S12-209 Effektivisering Av Energiförbrukning

Transcript

Skogsindustriella programmet Improved secondary heat systems and reduced water consumption within the pulp and paper industry Magnus Forslund, Åsa Samuelsson Effektivisering av energiförbrukning och minskad användning av processvatten Improved secondary heat systems and reduced water consumption within the pulp and paper industry Magnus Forslund, Marta Bialik, Karin Lindgren and Åsa Samuelsson S12-209 VÄRMEFORSK Serviceaktiebolag 101 53 STOCKHOLM ∙ Tel 08-677 25 80 Juni 2014 ISSN 1653-1248 VÄRMEFORSK Abstract The study presents process and cooling water consumption for some of the main products in the Swedish pulp and paper industry. Different process closure concepts and technology have been reviewed. Specifically increased bleach plant closure including effects on non-process elements (NPEs) has been simulated in computer models. A structuralized and hands-on method for analysing cooling and secondary heat has been established within the study. A number of the typical measures for increasing secondary heat recovery and increasing the utilization are described. An optimized sizing principle of heat recovery units and coolers is described. Key words: Water conservation, energy conservation, secondary heat, cooling systems, process closure, process simulation, bleach plant closure, process water, water cost, nonprocess elements (NPE). Vattenbesparing, energibesparing, sekundärvärme, kylsystem, slutning, processimulering, processvatten, blekerislutning, kostnad vatten, processfrämmande grundämnen (PFG). i VÄRMEFORSK Sammanfattning Denna rapport sammanfattar arbete som utförts inom Värmeforsk Skogsindustriella program, projekt nr S12-209, med extra finansiering från Å-Forsk, referens nr 11-179. Arbetet utfördes i samarbete mellan Innventia och ÅF Forest Industry som en skrivbordsstudie. Låg processvattenförbrukning och resulterande låga avloppsflöden är av väsentlig betydelse för energiförbrukningen vid produktion av papper- och massaprodukter. Målet inom Värmeforsks Skogsindustriella program är att minska vattenförbrukningen med 2% under programperioden 2013-2015. Den uppskattade besparingen inom Värmeforskprogrammet 2012-2014 motsvarar 2 TWh värme. Syfte Det övergripande målet inom studien är att minska process-och kylvattenanvändning och därmed också förbättra energieffektiviteten för tre huvudtyper av papper och massabruk. De tre anläggningstyperna är:  Produktion av helblekt avsalu sulftatbarrmassa  Integrerad produktion av oblekt kraftliner  Integrerad produktion av mekanisk massa och journalpapper I kvantitativa termer redovisas målet i denna studie för att minska energi och vattenförbrukningen enligt tabell nedan: Projektets mål för minskad energiförbrukning Total process energy, kWh/ADt whereof power kWh/ADt Helblekt avsalumassa Integrerad produktion av oblekt massa och liner Integrerad produktion av mekanisk massa och magasinpapper - 280 - 280 - 280 - 50 - 50 - 75 Ovanstående mål motsvarar en minskning med 2 200 GWh per år av processenergi, varav 570 GWh som el. Detta skulle innebära en minskning med drygt 2 % av den totala energianvändningen inom den svenska pappers- och massaindustrin. ii VÄRMEFORSK Inget mål kvantifieras för potentiella processvattenbesparingar i projektet. Dock är det allmänna målet för minskning vatten inom Värmeforsks Skogsindustriella program 2 % enligt ovan. Tidigare forskning och ett historiskt perspektiv En omfattande genomgång av tidigare forskning för ökad slutning av massa- och pappersbruk utgjorde den första delen av studien. Även en genomgång av alternativa utformningar av kylvattensystem och förbättringar av sekundärvärmesystem genomfördes. Under 1970, 1980 och början av 1990-talet var miljöpåverkan från skogsindustrin hög. Bruken hade inga externreningsanläggningar och energiförbrukningen var också högre. Normalt behövdes stödbränsle i form av brännolja för att ge processen tillräckligt med energi Från 1980-talet till ca 2000 minskade användningen av energi och vatten samtidigt som avloppsvolymen i en typisk massafabrik minskade avsevärt. Dessa åstadkommas genom en kombination av flera förbättringar i massaprocessen; minskad vattenförbrukningen i vedgården, ny kokteknik för kemisk massa utvecklades med lägre slutkappatal, införande av syrgasdelignifiering tillsammans med förbättrad tvätt, förbättrad spillhantering, införande av ECF och TCF-blekning, effektiv återanvändning av svartlutskondensat i olika positioner och recirkulation av blekeriet filtratet till bruna tvätten. Olika tekniker för avlägsnande av både organiska och oorganiska joner utvecklades och testades. Men det var införandet av externrening som minskade miljöpåverkan mest. På energisidan har minskad färskvattenanvändning och återvinning av sekundärvärme bidragit till den kraftiga energiminskningen. En betydande del av den minskade energiförbrukningen är även relaterad till effektivare avdunstning av svartlut till högre torrhalt, återvinning av ånga från mekaniska raffinaderier och ökad torrhalt på pappersmaskiner etc., vilket är något som inte är direkt relaterat till sekundärvärme-och kylsystem. Redovisad energi- och vattenförbrukning Den här studien presenterar process- och kylvattenanvändning för några av de viktigaste produkterna för den svenska massa-och pappersindustrin. Brukens förbrukning jämförs med EU - BAT BREF 2013 samt en hypotetisk referensfabrik som utvecklats av ÅF och Innventia inom tidigare forskningsprogram. Referensfabriken uppdaterades senast 2010 och även till viss del i detta arbete. Energiprestanda presenteras för befintliga bruk och extrema referenser som finns i världen. Nyckeldata presenteras i tabellen nedan för avsalubruk. I rapporten presenteras motsvarande uppgifter för kraftliner och magasinpapper. iii VÄRMEFORSK Jämförelse av referensfabrik, BREF BAT, svenskt genomsnitt och extremfall för helblekt sulfatmassabruk i världen. Processvattenavlopp dito COD Kylvattenavlopp dito värme m3/ton kg/ton m3/ton GJ/ton Referens bruk avsalu sulfatmassa 22 <9 fotnot2 7.6 Ångförbrukning Elförbrukning GJ/ton kWh/t 8.9 600 BREF BAT Svenskt genomsnitt 2012 Extrema bruk i världen 25-50 7-20 fotnot3 fotnot6 35 15.6 42 ≈ 5.3 23 21 fornot5 fornot6 13.7-18.4 700-800 13.9 796 9-10 520 (draft 2013) Utsläpp och förbrukning enligt BREF–BAT representerar en europeisk överenskommelse av bästa tillgängliga teknik inom massa- och pappersindustrin. Ovanstående siffror föreslås i det senaste utkastet daterad juli 2013. Dessa förväntas bli slutgiltigt bekräftade och antagna under 2014 [1]. Skillnaden mellan ett genomsnittligt svenskt bruk och referensfabriken är stor när det gäller vattenanvändning, 22 m3/ADt för referensfabriken och 35+42=77 m3/ADt för ett genomsnittligt svenskt avsalubruk. Den största delen av skillnaden beror på utformningen av kylvattensystemet som i referensfabriken är integrerad med processvattnet och innefattar kyltorn. Baserat på rapporterad vattenförbrukning bedöms svenska bruk ha processvattenförbrukningar och avloppsvolymer i linje med internationella normer och bästa tillgängliga teknik. Referensfabriken är en teoretisk greenfield fabrik som använder bästa tillgängliga beprövade teknik, där samtliga delar ska vara kommersiellt testade men inte nödvändigtvis i samma fabrik. Vattenförbrukningen i många befintliga fabriker är betydligt högre än i referensfabriken och en orsak till detta är att referensfabriken motsvarar en nybyggd (greenfield) fabrik. 1 Extremfall inkluderar tertiärrening vilket inte är jämförbart med referensfabriken, BAT och svensak bruk vilka normalt har sekundärrening. 2 Referensfabriken har kyltorn så inget kylvatten produceras 3 Ingen information tillgänglig iv VÄRMEFORSK Ytterligare processvatten slutning och optimering av sekundärvärmesystem För en fabrik som vill minska vattenanvändningen finns ett antal möjligheter, förbättrad integrering av kyl- och processvatten är ett sådant alternativ. Andra är att minimera spill, använda kondensatet från svartlutsindunstning i större omfattning och införa alkalisk slutning om det inte redan genomförts. Genom en screening har ett antal möjliga energi- och vattenbesparande åtgärder identifierats. Vissa åtgärder valdes ut för att utvärderas ekonomiskt. Dessa var       Ytterligare slutning av blekeriet i sulfatmassabruken Indunstning av avloppsvatten i pappersbruk med överskottsånga Förbättrad värmeåtervinning från imångor från lösartanken i sulfatmassabruk Maximera värmeåtervinning från processen genom optimerad dimensionering av värmeåtervinningssystem Minimerad färskvattenförbrukning genom ökad storlek på nya och befintliga kylare Installation av kyltorn för minimerad färskvattenförbrukningen Den största potentialen för minskning av processvattenanvändningen för referensbruket finns i blekeriet där ökad slutning resulterar i lägre färskvattenanvändning och minskad avloppsvolym. Betydande minskning färskvattenförbrukningen i blekeriet är möjligt om ökad slutningen genomförs, och det färskvatten som används i blekeriet ersätts av lutångkondensat. Ökad slutning av blekeriet är ofta begränsad av att hög nivå av organiskt material leder till ökad kemikalieförbrukning och anrikning av processfrämmande grundämnen som kan bilda utfällningar. Referensbruket för avsalumassa och andra moderna anläggningar för tidnings- och magasinpappersproduktion har ett ångöverskott. Detta skulle kunna användas för att ytterligare minska avloppsflödet. En stor potential har identifierats för att förbättra sekundärvärmesystemet för kemiska massabruk genom effektivare återvinning av värme från lösartankens imångor. För nya installationer, och ombyggnation av befintliga system, rekommenderas att utnyttjandet av värme från lösartanken studeras noga. En strukturerad metod för analys av sekundärvärme- och kylsystem introduceras i rapporten tillsammans med en rad möjliga åtgärder för att öka värmeåtervinningen eller att utnyttja sekundärvärmeöverskott på ett effektivare sätt. En metod för optimal dimensionering av värmeåtervinningsaggregat med avseende på värmebesparingar kontra kapitalkostnader presenteras i studien. En tumregel för dimensionering av värmeåtervinningsenheter presenteras. Potentialen i samband med rätt val av värmeväxlarstorlek, typ och konfiguration betraktas allmänt som hög för att förbättra befintliga system och utformning av ny utrustning. v VÄRMEFORSK Även en metod för dimensionering av processkylare med avseende på kylvattenförbrukningen och kapitalkostnad presenteras. Även för kylare presenteras en tumregel för dimensionering. Energibesparingen relaterad till att förbättra prestanda i kylare anses vara relativt liten jämfört med att optimera andra processer inom massaoch pappersindustrin. Brukets färskvattenförbrukning kan dock minskas till låg kostnad genom att öka prestanda eller storlek på installerade kylare. Detta alternativ bör prioriteras före installation av kyltorn. Resultat En minskning av den redan låga processvattenförbrukningen bör vara möjligt för ett kemiskt sulfatmassabruk. För referensbruket skulle volymen till externreningen kunna minskas från 14 m3/ADt till 9 m3/ADt innehållande 16,8 kg / ADt av COD . Risken för utfällning av kalciumoxalat i blekeriet kommer inte att öka. Men för att hålla klorid halten i vit- och svartlut på samma nivå som innan behöver en lakningsenhet för elfilteraska installeras (en kloridnjure). Driftskostnaden kommer att minskas med 6-9 Mkr/år och investeringskostnaden är i storleksordningen 40-50 miljoner kronor om ett askalakningssystem inkluderas för att behålla kloridhalten på konstant nivå. I rapporten diskuteras ytterligare slutning av blekeriet. För att minska avloppsflödet ytterligare behöver lutångkondensatet och bakvatten användas i andra positioner än i blekeriet. Dessa positioner utanför blekeriet är inte lätt att identifiera eftersom färskvatten användningen redan innan är låg. En ökad slutning av blekeriet skulle också kräva avlägsnande av processfrämmande grundämnen, främst kalcium. Ökade överbäringar av COD ökar behovet av kemikalier i blekeriet. Referensbruket för tidnings- och magasinpappersproduktion kan minska processavloppet med upp till 65 % genom indunstning av avlopp med överskottsånga. Total kapitalkostnaden beräknas till cirka 200 MSEK. Den specifika totala kostnaden med tanke på både kapital och drift beräknas till 12 kronor per sparad m3 beräknad över utrustningens skattade total livstid. Reducerad vattenförbrukning genom att installera kyltorn skulle resultera i en specifik kostnad på 0,2 kronor per sparad m3. Optimerad dimensionering av värmeväxlare för kylning skulle ha lägre specifik kostnad än att installera kyltorn. Den elkraftförbrukningen i samband med pumpning av kylvatten kan reduceras genom att optimera dimensioneringen av nya kylare och förbättra prestanda för befintliga enheter. Den totala effektförbrukningen i samband med pumpning av de rapporterade 300 miljoner m3 kylvatten [2] som används inom massa-och pappersindustrin beräknas vara mindre än 100 GWh. En minskning med 10 % genom att optimera kylare skulle ge en marginell energibesparing men signifikant lägre vattenförbrukning. Ökad processvattenslutning i blekeriet och andra fabriksavdelningar minskar nödvändigtvis inte elförbrukning eftersom ytterligare processutrustning introduceras i interna reningsteg. vi VÄRMEFORSK Rapporten visar att det finns en potential att öka produktionen av hetvatten (80ºC) från lösartankens imångor. Detta åstadkoms genom att dela upp värmeåtervinningsenheten i två steg, ett kylningssteg och ett andra värmeåtervinningsstegs vilket matas med halvvarmt hetvatten (65°C) vilket det oftast finns överskott av. Denna lösning är särskilt intressant att applicera på moderna sodapannor där lösartankens imångor vanligtvis används som förbränningsluft i sodapannan. Rapporten visar att referensbrukets sodapanna kan producera cirka 500 m3/h mer hetvatten med potential att ersätta 15 MW primärvärme för uppvärmning av internt hetvatten eller leverans till ett fjärrvärmenät med en återbetalningstid kortare än 1 år. Om det är möjligt att applicera tekniken på alla sodapannor i Sverige skulle detta motsvara en potential på 600 GWh per år. De allmänna riktlinjerna för dimensionering av värmeväxlare och rutiner/instrumentering för att följa upp deras prestanda har potential att förbättra graden av värmeåtervinning i de svenska massa- och pappersbruken. Den totala energiförbrukningen i samband med uppvärmning av processvatten i temperaturområdet 40 - 90ºC uppskattas till cirka 5 % av massa-och pappersbrukens energiförbrukning vilket motsvarar ca 3 TWh. En optimerad dimensionering av värmeväxlare och förbättrad övervakning och uppföljning beräknas minska energiförbrukningen upp till 10 %, det vill säga 300 GWh. Flera andra processkoncept för att förbättra sekundärvärmesystemet presenterats i rapporten. Genomförandet av dessa fabriksspecifika åtgärder har potential att spara upp till 3 TWh totalt relaterade till uppvärmning processvatten (se ovan). Slutsatser och diskussion Ökad slutning av massa- och pappersbruk minskar avloppsflödet och vattenförbrukningen och därmed också miljöpåverkan. I Sverige är minimering av volymen och föroreningar i processavloppen allmänhet en större drivkraft än minskad färskvattenvattenförbrukning. Det första steget mot ökad slutning är att implementera den bästa tillgängliga teknik som beskrivs för referensfabrikerna samt övriga publicerade rapporter. Det är dock inte alltid möjligt att direkt tillämpa dessa tekniker i befintliga massa-och pappersbruk. I verkligheten införs denna moderna teknik steg för steg i samband med investeringar i helt nya anläggningsdelar. Således, för befintliga bruk är resan mot referensbrukets prestanda lång och kräver tid. Dock är Värmeforsks målsättning att minska vattenförbrukningen med 2 % fullt möjligt för samtliga bruk och de tillhörande kostnader är hanterbara. En betydligt större minskning skulle kunna vara möjligt även för referensbruket med redan mycket låg vattenförbrukning. För helt nya anläggningar, eller efter implementering av den senaste tekniken i en process del av äldre bruk, krävs ytterligare innovativa processer för ökad slutningsgrad. I detta fall kan extrem slutning av det kemiska sulfatmassabrukets blekeriet motiveras vii VÄRMEFORSK utifrån minskade driftskostnader och lägre miljöbelastning. Historiskt sett finns det erfarenhet av minskad tillgänglighet vid drift av mycket slutna blekeriet. Problem med utfällning av oorganiska salter, hög kemikalieförbrukning och ökad nivå av klorid och i vit- och svartlut förväntas. Ny teknik såsom asklakning, ökad renhet för lutångkondensat och nya metoder för att avlägsna processfrämmande grundämnen ger nya förutsättningar för slutning av blekeriet för både TCF och ECF massakvalitéer. Ökad slutning till extrema nivåer kan också ha negativa effekter såsom ökad användning av el och primärvärme, kemikalieförbrukning etc. När alternativet med ökad vattenanvändning utvärderas ska den samlade miljöpåverkan beaktas för det specifika fallet. Referensbrukets färskvattenanvändning i samband med kylning av processen är redan extremt låg på grund av införandet av kyltorn. För befintliga fabriker visar studien att färskvattenförbrukningen kan minskas genom optimering av kylvattensystem. Det ekonomiska incitamentet för att minska färskvattenförbrukningen relaterad till processkylning är relativt lågt i Sverige. Generellt är endast några få anläggningar begränsade på tillgänglighet av färskvatten, och produktions- och användningskostnaden för mekaniskt behandlat kylvatten är mycket låg. Om mängden färskvatten är begränsande ska dimensionering av processkylare ses över som ett första steg eftersom det är en kostnadseffektiv åtgärd för att minska kylvattenmängd. Som ett andra steg för att minska färskvattenförbrukningen, eller om mottagaren är känslig för termisk belastning av recipienten i samband med utsläpp av kylvatten, kan kyltorn installeras. Sekundärvärmesystemen skiljer sig avsevärt från anläggning till anläggning, även inom samma produktkategori. Sekundärvärmesystem är komplexa, och ofta är det en tidskrävande aktivitet för att kartera och optimera dessa. Förbättrad återvinning av sekundärvärme går hand i hand med möjligheten att utnyttja värmeöverskottet. Ett sekundärvärmeöverskott är lika vanligt som ett underskott. Rapporten ger exempel på åtgärder med låg kapitalkostand som resulterar i energibesparingar av relativt högt värde. Åtgärder och optimering av sekundärvärmesystem med hög potential att ge en kort återbetalningstid, men i förhållande till den totala bruket primärvärmeförbrukning bedöms den potentiella energibesparingar inte vara större än några få procent (se ovan 3 TWh för Svensk Industri). Den potentiella energibesparingen relaterad till de åtgärder som diskuteras i rapporten är i samma storleksordning som målet, 2.2 TWh. Det har dock varit svårt att hitta processvatten- och sekundärvärmeåtgärder som kan minska elförbrukningen med 570 GWh. Detta utmanande mål kan jämföras med det svenska PFE initiativet (Programmet För Energieffektivisering) som har rapporterat en minskning med 1 450 GWh [3]. viii VÄRMEFORSK Rekommendationer och användning Denna rapport kan tjäna som grund för vidare studier och eventuellt genomförande av ökad blekerislutning för produktion av avsalumassa. Den sammanställda och redovisade statistiken för svenska massa-och pappersbruk samt nyckeltal för referensbruken är värdefull information när kylvatten och avloppsvolymer för en enskild anläggning ska värderas. Det är även viktigt att förstå att skillnader i specifik vattenförbrukning inom varje produktkategori variera beroende på olika produktionsprocesser och produktkvaliteten, inte bara på grund av skillnader i teknisk standard. Syftet med studien är även att rapporten fungera som en “handbok” för att förbättra och förstå de sekundärvärme- och kylvattensystem i massa- och pappersbruk. Ett antal både allmänna och specifika åtgärder för att förbättra systemen diskuteras. I rapporten presenteras också en metod för teknisk-ekonomisk dimensionering av nya värmeåtervinningsaggregat och kylare. ix VÄRMEFORSK Executive Summary This report summarizes work performed within the Forest Industrial Program, Värmeforsk, project No S12-209, with additional financing by Åforsk, ref No 11-179. The work was performed in co-operation between Innventia and ÅF Forest Industry as a desk-top study. Low process water consumption and effluent flows are of significant importance for the energy used when producing pulp and paper products. The target within the Värmeforsk Forest Industrial Program is to reduce the water consumption by 2% within the program execution 2013-2015. The estimated saving within the Värmeforsk program 2012-2014 is quantified as 2 TWh heat saving. Objectives Overall objectives are to reduce process and cooling water consumption thereby also improving the energy efficiency for three main types of pulp and paper mills. The three basic mills are:  Non-integrated softwood bleached kraft market pulp  Integrated production of unbleached kraftliner  Integrated production of mechanical pulp and magazine paper In quantitative terms, the objectives within this study for reducing the energy consumption are in table below Objectives within this study for reducing the energy Total process energy, kWh/ADt whereof power kWh/ADt Non-integrated bleached kraft market pulp Integrated production of unbleached kraft liner Integrated production of mechanical pulp and magazine paper - 280 - 280 - 280 - 50 - 50 - 75 The above objectives corresponds to a reduction of 2 200 GWh per year process energy, whereof 570 GWh as electric power. This would imply a reduction of slightly more than 2% of the total energy consumption within the Swedish pulp and paper industry. No objective was initially quantified for the potential process water savings related to the project. However, the general target for water reduction within Värmeforsk Forest Industrial program is 2% as outlined above. xi VÄRMEFORSK Earlier research and historic perspective A comprehensive review of earlier research for increased process closure of pulp and paper mills formed the initial part of the study. Also a screening of alternative designs of cooling water systems and secondary heat improvements was conducted. During 1970, ’80 and the beginning of ’90 the environmental impact was high. The mills did not have any external effluent treatment facilities, the energy consumption was high and normally fuel oil was needed to support the process with sufficient amount of energy. From 1980-ies to about 2000, the use of energy, water and the amount of effluent in a typical pulp mill decreased considerably. These could be accomplished by a combination of several approaches; decrease in woodyard water consumption, improved in pulping technologies to lower kappa number, introduction of oxygen delignification together with improved washing, improved spill management, partial or total elimination of chlorine-based bleaching chemicals, efficient reuse of evaporation condensate in different positions and recirculation of bleach plant filtrate to brownstock washing. Different techniques for removal of both organic and inorganic ions were developed and tested. With the introduction of more extended effluent treatment process the environmental impact was decreased dramatically. On the energy side the reduced fresh water consumption and recovery of secondary heat have contributed to decreased energy consumption. Also a significant part of the reduced energy consumption related to more efficient evaporation and combustion of liquor, recovery of vapour from mechanical refiners and improved press dryness on paper machines, etc. which is not directly related to the secondary heat and cooling systems. Reported energy and water consumption data The study presents actual process and cooling water consumption for some of the main products in the Swedish pulp and paper industry. The performance of existing mills is compared with EU-BAT BREF 2013 as well as a hypothetical reference mill developed by ÅF and Innventia within several earlier research programs, updated in its latest revision 2010 and also to some extent within this work. Similarly energy performance data has been presented for existing mills and the extreme references available in the world. Key performance data is presented in table below for the market kraft pulp mill. The report present corresponding data for the kraftliner and magazine paper. xii VÄRMEFORSK Comparison of hypothetical reference mill, BREF BAT, Swedish average and lowest in operation for the market pulp mill. Reference Market pulp mill BREF BAT (draft 2013) Swedish Average 2012 Process water effluent ditto COD Cooling water effluent ditto heat load m3/ton kg/ton m3/ton GJ/ton 22 <9 note5 7.6 25-50 7-20 note6 note6 Steam heat consumption Power consumption GJ/ton kWh/ton 8.9 600 13.7-18.4 13.9 700-800 796 35 15.6 42 ≈ 5.3 Lowest in operation at extreme conditions 23 24 note5 note6 9-10 520 The BREF-BAT emissions and consumption numbers represents European agreement of the best-available-technology within the pulp and paper industry. The above guidelines are proposed in the latest draft dated July 2013, which also is foreseen to be accepted during 2014. The difference between an average Swedish mill and the reference market pulp mill is large when it comes to water use, 22 m3/ADt for the reference mill and 35+42=77 m3/ADt for the average Swedish mill. The main part of the difference is related to cooling water system design. Based on reported water consumption, we still consider that Swedish mills have process water consumptions and process effluent flows in line with international standards and BAT. The reference mills are theoretical greenfield mills using best available proven technology, but not necessarily commercially possible to justify for all mill conditions. Still it should be noted that the water consumption in many existing mills still is significantly higher than in the developed reference mills and one reason for this is the reference mill being a greenfield mill. 4 Extreme case incudes tertiary treatment stage not comparable to secondary treatment in reference mill, BAT and existing Swedish mills. 5 Reference mill design includes cooling towers resulting in no cooling water effluent. 6 Information not available xiii VÄRMEFORSK Further process water closure and optimization of secondary heat systems For a mill that would like to decrease the water use there is thus a number of opportunities. The integration of cooling water and process water is one such option. Others are minimising the spills, use the condensate from black liquor evaporation more extensively and implement alkaline closure if not already implemented. In a screening process a number of possible energy and water conservation measures and their potential were described. Some measures were selected to be evaluated further economically. These were:       Further closure of bleach plant in the kraft pulp mills Evaporation of effluent in paper mills with surplus steam Improved heat recovery from dissolving tank vent gases in chemical pulp mills Maximize heat recovery from process by optimized sizing of heat recovery units Minimized fresh water consumption by increase size of new and existing coolers Installation of cooling towers for minimized fresh water consumption The highest process water saving potential for the reference mill has been identified as further closure of the bleach plant resulting in lower fresh water consumption and effluent generation. Significantly decreased in fresh water in the bleach plant is possible if increased closure is applied and the fresh water used in the bleach plant is replaced by condensate. Increased closure of bleach plant is often restricted by high level of organic material leading to increased chemical consumption and non-process elements. The magazine and news print reference mill and very modern mills have a surplus of steam. This could be used for evaporation of the effluent and further reduction of the effluent discharge. A high potential has been identified for improving the secondary heat systems in chemical pulp mills through more efficient recovery of heat from the dissolving tank vent gases. For new installations and rebuilt of existing systems, it is recommended that the surplus heat utilization is carefully studied. A structuralized method for analysing secondary heat and cooling system is introduced together with a number of potential measures for increasing heat recovery or utilizing a secondary heat surplus more efficiently. A method for optimal sizing of heat recovery units with respect to heat savings versus capital cost is presented within the study. A rule of thumb for sizing is presented. The potential related to proper selection of heat exchanger size, type and configuration is generally regarded as high for improving existing systems and designing new equipment. xiv VÄRMEFORSK Also a method for selecting a suitable sizing for process coolers with respect to cooling water consumption and capital cost is presented. Also for coolers a rule of thumb for sizing is presented. The energy saving related to optimizing the performance of coolers is regarded as relatively small compared to optimizing other process in the pulp and paper industry. The mill fresh water consumption can however be reduced at low cost by increasing the performance or size of installed coolers. This option should be prioritized before installing cooling towers. Results A very low bleach plant effluent volume should be possible for a market kraft pulp mill. For the reference mill the effluent volume to external treatment could be decreased from 14 m3/ADt to 9 m3/ADt containing 16.8 kg/ADt of COD. The risk for scaling of calcium oxalate in the bleach plant will not increase. But to keep the chloride concentration in the white and black liquor on the same level as before a precipitator dust leaching unit should be implemented. The operating cost will be reduced by 6-9 MSEK/year and the investment cost is in the order of 40-50 MSEK if an ash leaching system is included maintain the same chloride level. In the report further closure of the bleach plant was discussed. In order to decrease the effluent further the condensate and white water need to be used in other positions than the bleach plant. These positions are not easy to find as the fresh water use is already low. An increased closure of the bleach plant would also require removal of non-process elements, mainly calcium. Increased carry-over of COD increases the chemicals needed in the bleach plant. The reference magazine paper mill effluent can be reduced by up to 65% through evaporation with surplus steam. Total capital cost is estimated to be around 200 MSEK. The specific total cost considering both capital and operation is calculated to be 12 SEK per reduced m3 of the effluent discharge. Reduced cooling water consumption by installing cooling towers would result in a specific cost of 0.2 SEK per saved m3. Optimizing the sizing of heat exchangers for cooling would have lower specific cost than installing cooling towers. The electric power consumption related to pumping cooling water can be reduced by optimizing the sizing of new coolers and improve the performance of existing units. The total power consumption related to pumping the reported 300 million m3 cooling water [2] used within the pulp and paper industry is estimated to be less than 100 GWh. A reduction for 10% by optimizing coolers would give a marginal power saving but significant lower water consumption. Reducing process water used directly in the process does not necessarily reduce the electric power consumption due to introduction additional process equipment for internal treatment stages. The report shows that there is a potential to increase the generation of hot water (80ºC) from the dissolving tank vent gas. This is accomplished by splitting the condenser in two stages, one cooling stage and a second heat recovery stage feed with a semi hot xv VÄRMEFORSK water (65ºC). This solution is specifically interesting to apply on modern recovery boilers where the dissolving vent gas usually is sent to the recovery boiler. The report indicates that the reference mill recovery boiler can produce around 500 m3/h more hot water with potential to replace 15 MW primary heat for water heating or usage for delivery to a district heating network with a pay-back less than 1 year. If possible to apply on all the recovery boilers in Sweden this would correspond to 600 GWh per year. The general guidelines for sizing heat exchangers and analyzing their performance have potential to improve the degree of heat recovery in the Swedish existing mills. The total energy consumption related to heating process water in the temperature range 40-90ºC is estimated to be around 5% of the pulp and paper energy consumption corresponding to approximately 3 TWh. A proper sizing of heat exchangers and improved monitoring and follow-up is estimated to reduce this energy consumption up to 10%, i.e. 300 GWh. Several other process concepts for improving the secondary heat system has been presented in the report. Implementation of these rather mill specific measures has potential of saving up to the estimated total of up to 3 TWh estimated to be related to heating process water above. Conculsions Increasing the closure of pulp and paper mills reduces the effluent emissions and the water consumption, and thereby also the environmental impact. In Sweden the effluent emissions is generally a higher concern compared to the fresh water consumption. The first step towards increased closure is to implement the best available technology described in the reference mills. However, it is not possible to directly implement this energy and water efficient technology in existing pulp and paper mills. In reality this modern technology is introduced step-by-step in conjunction with replacement of the expensive main equipment. Accordingly, for existing mills the step towards the reference mill performance requires time. However, to reach the goal by Värmeforsk, to decrease the water consumption by 2%, should be possible for all mills and the related costs are manageable. Significantly larger decrease could be possible for the reference mill with very low water consumption. For completely new mills or after implementing the latest technology in a process section of older mills, further innovative process closure concepts are required. In this case, increased closure of the market pulp bleach plant may be justified based on reduced operating costs and lower environmental performance. Historically, there is experience of reduced availability when operating with very closed bleach plant. Problems with scaling of inorganic salts, high chemical consumptions and increased level of chloride in white and black liquor is expected. New technology such as ash leaching, improved evaporation plant condensate quality and novel methods to remove non-process elements may enable new preconditions for closure of the bleach plant for both TCF and ECF pulp qualities. xvi VÄRMEFORSK Increasing the closure to extreme levels can also have negative impacts such as increased use of heat and power, chemical consumption etc. When evaluating the overall environmental impact also this needs to be considered for the specific case. The reference mill fresh water usage related to cooling purposes is already extremely low due to the introduction of cooling towers. For existing mills the study shows that the fresh water consumption can be reduced by optimizing the cooling water systems. However, the economic incentive for reducing the fresh water consumption related to cooling purposes is relatively low in Sweden. Generally, few mills are limited on the fresh water availability and the production and usage cost of mechanically treated cooling water is very low. If the quantity of fresh water is limited, the sizing of process coolers should be reviewed as a first low cost step in order to reduce the cooling water quantity. As a second step for reducing consumption, or if the recipient is sensitive for thermal load related to the discharge of cooling water, cooling towers can be installed. The secondary heat systems differ considerable from mill to mill, also within the same product category. Secondary heat systems are complex and often it is a time consuming activity to audit and describe them. Improved recovery of secondary heat goes hand-inhand with the possibly of finding the utilization of the heat. An excess of secondary heat is as common as a deficit of secondary heat. The report gives examples of low capital cost measures that results in relatively high value energy savings. Optimization of secondary heat systems has a high potential for short pay-back measures but relative to the total mill primary heat consumption the potential energy saving if improved process integration is expected to not be higher than a few percent. The potential total energy saving related to the measures discussed in the report is in the same order of magnitude as the objective. However, it has been difficult to find process water and secondary heat measures that can reduce electric power consumption by 570 GWh per year. This challenging target can be compared to the Swedish PFE-program (programmet för energieffektivisering) which has reported a reduction of 1 450 GWh per year [3]. xvii VÄRMEFORSK Recommendations and use The study report can serve as basis for further study and possible realization of bleach plant closure concepts for the market pulp mills. The presented water consumption statistics for Swedish pulp and paper mills together with the reference mills are valuable information when benchmarking cooling water and process effluent flows of a real mill. It is although important to understand that differences in specific water consumption within each product category can vary due to differences in the production process and product quality, not only because of differences in technical standard. The aim of the study is to also serve as a “handbook” for improving and understanding the secondary heat and cooling systems in pulp and paper mills. A number of both general and specific measures for improving the systems are discussed. The report also presents a method for techno-economic design of new heat recovery units and coolers. xviii VÄRMEFORSK Table of contents 1 INTRODUCTION ................................................................................................ 1 1.1 1.2 1.3 1.4 2 BACKGROUND ...................................................................................................... 1 EARLIER RESEARCH ............................................................................................... 2 OBJECTIVES......................................................................................................... 4 REPORT STRUCTURE.............................................................................................. 5 ENERGY AND WATER CONSUMPTION DATA ..................................................... 7 2.1 DEFINITIONS ........................................................................................................ 7 2.1.1 2.1.2 Process and cooling water ......................................................................... 7 Heat and power consumption ..................................................................... 8 2.2 KEY FEATURES A WELL-DESIGNED PROCESS WATER AND SECONDARY HEAT SYSTEM .......... 8 2.3 REPORTED CONSUMPTION DATA IN SWEDISH MILLS ................................................... 10 2.4 ENERGY AND WATER CONSUMPTION IN THE REFERENCE PULP MILL ............................... 12 3 PROCESS CLOSURE OF THE BLEACH MARKET PULP MILL ............................. 14 3.1 OVERVIEW ......................................................................................................... 14 3.2 FURTHER CLOSURE OF THE REFERENCE MILL BLEACH PLANT ....................................... 14 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 Reference mill .......................................................................................... 14 Decreased water use by increased filtrate use ......................................... 15 Results .................................................................................................... 17 Internal dilution before D0 ........................................................................ 20 Additional possibilities. ............................................................................ 21 3.3 RETURN ACID FILTRATES FROM BLEACHING TO LIQUOR CYCLE ..................................... 21 3.4 SEPARATE EVAPORATION AND INCINERATION OF ACID FILTRATES FROM BLEACHING ........ 22 3.5 PARTIAL RECOVERY OF PROCESS EFFLUENT TO WATER TREATMENT PLANT OR SELECTED POSITIONS IN THE PROCESS .......................................................................................... 22 3.5.1 3.5.2 3.6 3.7 3.8 3.9 3.10 3.11 4 MEMBRANE FILTRATION OF EFFLUENTS AND INTERNAL REUSE ..................................... 23 ALTERNATIVE USAGE OF EVAPORATION CONDENSATES .............................................. 24 REDUCED CONSUMPTION AND RECOVERY OF SEALING WATER ..................................... 26 RECOVERY OF PURGE FROM COOLING TOWERS ......................................................... 27 COLLECTION OF STEAM CONDENSATE FROM PIPE RACKS ............................................ 27 RETURNING COOLING WATER FROM OIL COOLERS ...................................................... 28 PROCESS CLOSURE OF THE KRAFTLINER PULP MILL .................................... 29 4.1 4.2 4.3 4.4 4.5 5 Recovery to water treatment plant ........................................................... 23 Recovery directly to process .................................................................... 23 OVERVIEW ......................................................................................................... 29 MEASURES FOR THE MARKET PULP MILL THAT ARE VALID FOR KRAFTLINER ..................... 29 COMPLETE EVAPORATOR CONDENSATE REUSE ......................................................... 29 USING PAPER MACHINE WHITE WATER FOR PULP WASHING .......................................... 30 MEMBRANE FILTRATION OF EFFLUENTS AND INTERNAL REUSE ..................................... 30 PROCESS CLOSURE OF THE MAGAZINE PAPER MILL ..................................... 31 5.1 MEMBRANE FILTRATION OF EFFLUENTS AND INTERNAL REUSE ..................................... 32 5.2 RECOVERY OF EFFLUENT TO WATER TREATMENT OR SELECTED POSITIONS IN THE PROCESS32 5.3 EVAPORATION OF EFFLUENT FROM TMP PLANT......................................................... 33 6 OPTIMIZED SECONDARY HEAT AND COOLING SYSTEM .................................. 34 6.1 INTRODUCTION ................................................................................................... 34 6.1.1 6.1.2 6.1.3 Objectives of secondary heat and cooling water systems ......................... 34 Principles of efficient heat recovery ......................................................... 35 Definition of coolers and heat recovery units ............................................ 36 xix VÄRMEFORSK 6.2 DESIGN TO IMPROVE SECONDARY HEAT SYSTEM IN MILLS ............................................ 37 6.2.1 6.2.2 6.2.3 Methodology ............................................................................................ 37 Improving heat recovery and minimize consumption of secondary heat .... 39 Utilization of high temperature secondary heat ........................................ 50 6.3 ALTERNATIVES FOR REDUCED COOLING WATER CONSUMPTION AND RELATED ENERGY CONSUMPTION ........................................................................................................... 56 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5 6.3.6 7 Principles to minimize cooling water consumption .................................... 56 Primary heat consumption ....................................................................... 56 Avoid recovery of low grade heat from vapours ........................................ 57 Minimize cooling water flow by increasing delta T..................................... 57 Cooling by delivery of secondary heat to external consumers ................... 58 Cooling water system with cooling towers ................................................ 59 EVALUATION OF SELECTED ALTERNATIVES .................................................. 60 7.1 SELECTION OF ENERGY AND WATER CONSERVATION MEASURES ................................... 60 7.2 EVALUATION OF SELECTED PROCESS WATER CONSERVATION OPTIONS ......................... 65 7.2.1 7.2.2 Increased closure of the reference mill bleach plant................................. 65 Evaporation of process effluent in the integrated magazine paper mill ...... 66 7.3 EVALUATION OF SELECTED SECONDARY HEAT IMPROVEMENTS AND COOLING WATER CONSERVATION OPTIONS.............................................................................................. 68 7.3.1 7.3.2 7.3.3 7.3.4 7.3.5 8 Two stage heat recovery of dissolving tank vent gas ................................ 68 Installing cooling towers .......................................................................... 69 Optimal sizing of new heat recovery units ................................................. 70 Optimal sizing of new coolers ................................................................... 72 Increase size and number of existing coolers ........................................... 74 RESULTS AND CONCLUSIONS ........................................................................ 75 8.1 COMPILED WATER AND ENERGY CONSUMPTION DATA ................................................. 75 8.1.1 8.1.2 8.1.3 Market pulp .............................................................................................. 76 Kraftliner ................................................................................................. 76 Magazine Paper ....................................................................................... 77 8.2 IMPROVEMENT MEASURES ..................................................................................... 78 8.2.1 Measures for reducing the fresh water consumption ................................ 78 8.2.2 Measures for reducing the process effluent flow and environmental impact80 8.2.3 Measures for optimizing heat recovery and utilization in secondary heat systems ............................................................................................................... 81 9 RECOMMENDATIONS AND USE ....................................................................... 83 10 FUTURE STUDIES ........................................................................................... 84 11 LIST OF ABBREVIATIONS................................................................................ 85 12 REFERENCES ................................................................................................. 86 xx VÄRMEFORSK Appendices A EARLIER RESEARCH AND TRENDS B REPORTED ENERGY AND WATER CONSUMPTION DATA C DESCRIPTION BLEACHED MARKET PULP – REFERENCE MILL D DESCRIPTION KRAFTLINER – REFERENCE MILL E DESCRIPTION MAGAZINE PAPER – REFERENCE MILL F MASS BALANCES AND BLOCK DIAGRAMS G ENERGY BALANCES H WATER BALANCES I OPTIMAL DESIGN OF HEAT RECOVERY UNITS J OPTIMAL SIZING OF COOLERS K DRIVERS FOR REDUCED PROCESS WATER CONSUMPTION AND IMPROVED SECONDARY HEAT SYSTEMS L KEY DATA FOR BLEACH PLANT CLOSURE xxi VÄRMEFORSK 1 Introduction 1.1 Background This report summarizes work performed within the Forest Industrial Program, Värmeforsk, project No S12-209, with additional financing by Åforsk, ref No 11-179. The financial contribution from the different parties is summarized below: Table 1.1: Financial contribution Financial contribution to project No S12-209 Värmeforsk Åforsk In kind ÅF/ Innventia Total SEK 350 000 350 000 100 000 800 000 The work was performed in co-operation between Innventia and ÅF Forest Industry as a desk-top study. Basic data and information was collected from Innventia and ÅF file data and public sources. The primary objective for the Forest Industrial Program within Värmeforsk is to improve the energy efficiency of the pulp and paper production. Higher energy efficiency of the processes would make more energy available for production of power or alternative biorefinery products, or alternatively improve the degree of selfsufficiency in the industry. Increased availability of biomass, biobased by-products, and increased production of renewable power as well as delivery of district heating to communities will improve chances of fulfilling the national environmental targets. Low process water consumption and effluent flows are of significant importance for the energy consumption when producing pulp and paper products. The target within the Forest Industrial Program is to reduce the water consumption by 2%. The estimated saving is quantified as 2 TWh heat saving. In addition to this comes savings related to chemicals and electric power which as the same order of magnitude in economic values. The aim of this study is to explore new technology and process concepts that can reduce the water consumption as far as possible for the Swedish pulp and paper mills. Methods of achieving low consumption numbers related to water usage are developed within the project and have been presented for representatives in the industry. The required investment and operation costs related to implementing new technology must be considered when targeting lower process water consumption. Also the presents and built-up of NPE:s (Non-Process-Elements) needs to be addressed with further closure of mills as well as the increased consumption of bleaching chemicals due to increased circulation of organic substances. 1 VÄRMEFORSK 1.2 Earlier research Historically, a pulp mill or an integrated pulp and paper mill were counted among industry’s largest fresh water consumers and most serious water polluters. Also significant amount of fossil fuels where used. The production process was not resource efficient and did not include systems for recovery of secondary heat. The process effluent contained high concentrations of organic material (BOD and COD) as well as chlorinated organic compounds, AOX (adsorbable organic halogens), which lead to serious disruptions of the aquatic life in the recipient and was characterized by high acute toxicity and/or high bioaccumulation. As the older mills during 1970, ’80 and the beginning of ’90 did not have any external effluent treatment facilities, decreasing the effluent volume and its environmental load by internal cleaning and/or recirculation was their only option for decreasing the total environmental impact of the effluents. A report by National Council for Air and Stream Improvement [4] offers an interesting description of the industrial trends in process closure as well as a summary of the developed technical solutions and mill case studies. In conjunction with the development in higher degree of process closure there has been a development towards higher self-sufficiency of energy. Reduced fresh water consumption and recovery of secondary heat has contributed to lower energy consumption. Figure 1 below shows the development for the Swedish pulp and paper industry regarding some selected energy indicators. Figure 1: Development of energy indicators for Swedish pulp and paper industry. Also a significant part of the reduced energy consumption related to more efficient evaporation, recovery of vapour from mechanical refiners and improved press dryness 2 VÄRMEFORSK on paper machines, etc. which is not directly related to the secondary heat and cooling systems. In short, the mills aimed at gradually adopting a holistic approach to waterborne pollution minimization by: prevention of the generation of large effluent volumes with undesired compounds, reuse of the internal water resources and recovery of the organic compounds from water streams. The above could be accomplished by a combination of several approaches:  Decrease in woodyard water consumption  Improved in pulping technologies to lower kappa number with maintained yield.  Introduction of oxygen delignification together with optimized washing in cooking, brownstock washing bleach plant.  Improved spill management.  Partial or total elimination of chlorine-based bleaching chemicals.  Efficient reuse of evaporation condensate in different positions instead of fresh water.  Recirculation of bleach plant filtrate to brownstock washing and other process areas. Recirculation of internal streams with high content of organics and non-process elements brought a variety of process problems and therefore techniques for removal of both organic and inorganic ions were developed and tested. Some of the techniques for bleach plant filtrates are listed below.  Flocculation/precipitation  Membrane filtration  Evaporation  Ozone treatment  Ion exchange But with the development of the environmental science and higher energy cost, including new tools for studying the environmental performance of the industrial sites as well as implementation of external treatment, the attitude towards the mill closure and self-sustainability on energy has changed. The main conclusion from Alliance for Environmental Technologies panel 1997 [4] stated that a total elimination of bleaching effluents should not be given highest priority. An extensive, or even full, bleach plant closure may even be undesirable if the consumption of raw materials, energy or other resources increases significantly. From 1980-ies to about 2000, the use of energy, water and the amount of effluent in a typical pulp mill decreased considerably, and although the raw effluent contaminant levels did so only slightly, the total contaminant discharge from the treated effluents decreased much more. The reduction in water usage was caused probably more by implementing effluent treatment processes rather than the actual closure, although some 3 VÄRMEFORSK process modifications were also involved. Further closure would certainly bring about even more profound reduction in effluent load; it would however lead to some extent an increased consumption of energy and chemicals. Only if particularly strict emission limits or water savings are to be met is a high degree of closure recommended. For the above reasons, the interest in minimizing the flow of effluents from the mill by increased process closure and internal purification techniques has gradually declined during the end of 1990’ and the first decade of 21st century. But during the last years there is a renewed interest for increasing the degree of mill closure and there are three main driving forces behind this phenomenon.  Many mills operate within ever tighter economic margins and increasing the production is seen as an easy possibility to increase the revenue. However, the production increase is difficult to achieve within the current emission limits. It is also expected that the future emission permits may become stricter.  Increasingly larger amounts of pulp and paper goods are being produced in localisations with limited access to fresh water, or even temporary or permanent water shortages.  There is a growing interest among customers to buy sustainable pulp and paper products with a high environmental profile. One of the important parameters when evaluating the product’s environmental impact is fresh water consumption during the production stage, often expressed as water footprint. A comprehensive review of earlier research and future trends is provided in Appendix A together with closure activities at different pulp and paper mills. The interest for increased self-sustainability of energy has been grown continuously since the 1970-ies. During the first decade of 21st century the focus has although changed towards biorefinery processes for production of new materials and biofuels. Introduction of the new biorefinery processes will be a major driver for increasing the energy efficiency of existing pulp and paper processes and the related secondary heat and cooling systems. 1.3 Objectives Overall objectives are to reduce process and cooling water consumption thereby also improving the energy efficiency for three main types of pulp and paper mills. The three basic mills are:  Non-integrated softwood bleached kraft market pulp  Integrated production of unbleached kraftliner  Integrated production of mechanical pulp and magazine paper In quantitative terms, the objectives within this study for reducing the water consumption and its associated energy conservation are: 4 VÄRMEFORSK Table 1.2: Objectives within this study for reducing the energy and water consumption Total process energy, kWh/ADt whereof power kWh/ADt Non-integrated bleached kraft market pulp Integrated production of unbleached kraft pulp and liner Integrated production of mechanical pulp and magazine paper - 280 - 280 - 280 - 50 - 50 - 75 The above objectives corresponds to a reduction of 1 600 GWh per year process energy, whereof 325 GWh as electric power, if limited to the product quantity of above selected pulp and paper products. This would imply a reduction of 2% of the total energy consumption within the Swedish pulp and paper industry. No objective was initially quantified for the potential process water savings related to the project. However, the general target for water reduction within Värmeforsk Forest Industrial program is 2%. The comparison is made with the reference mills developed in within the FRAM program and updated 2010 by ÅF and Innventia [5] [6]. Rough assessment of investment and operating costs for selected alternatives are provided. 1.4 Report structure Chapter 2: Data on specific energy and water consumption for actual operating mills is provided including data for the hypothetical reference mills. Chapter 3-5: Measures for reducing process water consumption for the three basic mill types are presented including potential technologies that could be considered in order to achieve very high degrees of water conservation. Chapter 6: Discussion on potential measures and technologies in secondary heat and cooling systems for achieving high energy efficiency and low water consumption. Chapter 7: Technical and economical assessment of selected alternatives regarding water and heat conservation for the three basic mill types. 5 VÄRMEFORSK Chapter 8: Results as well as analysis of water and energy data when comparing average of actual mills, “best performing actual mill”, BAT and reference mills. Chapter 9-11: Conclusions, recommendations and suggestions for future studies. Chapter 12: References Appendices A-L: Reference data and information relevant for some of the chapters. The reference group from Värmeforsk consist of the following persons:       Kristian Elvemo, SCA Munksund Mattias Eriksson, SCA Östrand Bertil Lundberg, BillerudKorsnäs Skärblacka Kajsa Fougner, Stora Enso Skoghall Torsten Svenland, Stora Enso Nymölla Thomas Eriksson, Pöyry 6 VÄRMEFORSK 2 Energy and water consumption data 2.1 Definitions 2.1.1 Process and cooling water The water consumption within the pulp and paper industry can be divided into two main categories:  Process water: Water, which is used directly in the process for washing, controlling of pulp consistency, make-up in boiler water production, make-up in the liquor cycle, chemical preparation, wood yard irrigation, debarking etc. The process water leaves the system via the effluent treatment plant.  Cooling water: Water where the main purposes is to remove heat from the process. Heat leaves the system either as warm cooling water going to the recipient or as vaporization heat in the cooling tower fume. The cooling water usually leaves the system via an arrangement separated from the process effluent. The process water system is normally separated from the cooling water system. The separation starts at the raw water treatment plant where the water often is divided into different qualities (mechanically filtered, chemically treated etc.). Mechanically filtered water is normally used for cooling purposes. The process consumes a variety of chemically and mechanically treated water. The systems for the process water and the cooling water are normally heat integrated. For example, a significant amount of process heat leaves the system with the process effluent. The treated process effluent and the warm cooling water leave the mill. When comparing the raw water consumption with the total effluent volume (cooling water + process water) there is generally only a very small difference. For market pulp mills only 1-2 m3/ADt leaves the overall system as vapour resulting in a slightly lower effluent flow compared to raw water consumption. The difference is also negligible at the production of other pulp and paper products. This means that it is practically possible to analyse the mill raw water consumption by measuring the effluent flow. The effluent flow is more often monitored with higher precision due to continuous follow-up of environmental permits. However, note that in mills without recycle of the cooling water through cooling towers, the cooling water effluent flow needs to be measured separately if not discharged together with the process effluent flow. 7 VÄRMEFORSK 2.1.2 Heat and power consumption Energy consumption is a well-used expression. The term energy consumption rather refers to a transfer of energy. We transfer high value energy and release it to the surroundings as more or less low value heat. The heat in steam and steam condensates in the mill is called primary heat. Primary heat also includes fuels such as fuel oil, natural gas, wood residuals etc. which are used in direct-fired units such as lime kilns. Electric energy is also primary heat. The primary energy consumption of the production process is often reported for into two categories:   Process steam heat consumption: This energy consumption includes all heat delivered to the process through the steam distribution and is normally expressed as GJ per ton product. Heat used for cogeneration of electric power is not included in the process steam heat consumption; nether delivery of steam to external process not related to the production process such as for example district heating or integrated saw mills. The origin of the steam heat is the fuel used in the steam boiler. Process power consumption: This energy consumption includes all electric power consumed for producing the product. Electric power used in boiler for steam heat generation is normally not included. This figure refers to the gross electric power consumption excluding possible internal cogeneration of power in turbines. Primary heat is consumed in the production process and the heat that can be recovered is called secondary heat. The secondary heat system and cooling systems starts where the primary heat is transferred to the process. The primary heat, secondary heat and cooling systems are all therefore interconnected. There are however no established accounting principles for secondary as is the case for primary heat. Low process steam heat consumption is although both the target and result of high recovery of secondary heat 2.2 Key features a well-designed process water and secondary heat system Foremost, a well-designed process water and secondary heat system is characterized by low process water and cooling water effluent flow combined with a low primary heat consumption. The process effluent volume that enters the effluent treatment is low, and pollutants are concentrated which facilitates the effluent management. In real operation an optimized secondary system is recognized by negligible steam heat consumption used for producing hot/warm water. Also low steam consumption in bleaching process and white water system around the drying/paper machine indicates that high water temperatures can be maintained by recovery of secondary heat. Furthermore, the produced make-up water to the boiler should be preheated to around 80ºC with secondary heat in a well-functioning system. 8 VÄRMEFORSK A surplus of high temperature secondary heat is also expected in a good secondary heat system. This surplus heat should be available at around 80-90ºC for chemical pulp mills, but lower for integrated, and specifically in non-integrated paper mills where the secondary heat sources generally have a lower temperature. The surplus heat is normally delivered to a nearby district heating net or any other external process. The high temperature heat leaves the process with the cooling water if there is no useful utilization available. In existing Swedish pulp and paper mills a good cooling water system is characterized by generously sized coolers that can keep down the volume flow, and at the same time keep up the exiting cooling water temperature. Also proper segregation of cooling water avoiding it to be mixed with process effluent is desirable. In mills with a long history of rebuilds and modifications has made the cooling systems difficult to survey resulting in mixing of process and cooling water. Generally, completely new pulp and paper mills including the latest technology have a lower cooling requirement. This manly because the of an overall lower primary heat consumption which indirectly results in lower cooling due when the heat leaves the system. Cooling towers are only required for new mills located in areas where the available fresh water is limited. The key principles of a good secondary heat design and water management are the same for improving existing mills as well as designing new greenfield mills. Important to understand is that the difference in age/technology of the expensive main equipment results in higher consumption numbers for existing mills despite an optimized process. (see also discussion in chapter 8) Typically low process water consumption is primarily achieved by introducing efficient washing technology for pulp combined with extensive countercurrent washing without mixing alkaline and acidic filtrates. The selected washing equipment has a large impact on the water use in the bleach plant, wash presses are preferred when it comes to low water use. Evaporation condensates and drying machine white water are reused in kraft pulping as far as possible considering product quality. Similarly integrated mechanical pulp and paper mills reuse the white water from the paper machine. Fresh water is primarily used in the final processing stages where the water quality requirement is highest. Spills should be kept down and reused in the process as far as possible. Alkaline closure is applied. The steam heat consumption related to heating of process water is minimized by process integration where the general principle is to cool hot streams with hot media and colder streams with colder media. The cooling and heating is performed in generously sized heat exchangers reducing the temperature difference between but still within what can be economically justified. 9 VÄRMEFORSK Methodology for analysing and options for improving secondary heat and cooling systems is presented in chapter 6. Alternative concepts and technology for reducing the process water consumption are discussed in chapter 3, 4 and 5. 2.3 Reported consumption data in Swedish mills Energy and water consumption data has been collected within this study for the selected products, i.e. market pulp, kraftliner and newspaper/magazine paper. A more detailed assessment and analysis of data is enclosed in Appendix 0. Statistics regarding effluent flows, and corresponding raw water consumptions, from the Swedish pulp and paper industry has been collected from Skogsindustrierna. The information is available electronically on the website from year 2001 to 2012 [2]. The specific numbers shown in the tables in this chapter have been calculated based on annually produced pulp and paper quantities reported by the mills. Historic data is available in reports issued by NV (Swedish EPA) not published on the website [7]. In this context market pulp mills refer to mills producing primarily bleached (ECF and TCF) paper grade kraft pulp for the open market. These mills are not integrated with paper mills, i.e. all pulp is sold. The effluent volume, and corresponding raw water consumption, for the Swedish kraftliner mills are expressed per ton of liner. The mills use various amounts of fibres for recycling in their liner products, 20-45 %. Effluent volumes in Swedish newspaper and magazine paper mills are expressed per ton of produced paper. Also in these mills the amount of fibers for recycling varies. Table 2.1: Summary of average process and cooling water effluent for the Swedish mills. Market pulp Process effluent year 2012 (ditto year 2000) (ditto year 1990) Cooling water effluent year 2012 m3/ton m3/ton m3/ton m3/ton 35 (58) (108) 42 Newspaper and magazine 14 (14) (19) 28 Total effluent year 2012 m3/ton 77 42 Kraftliner 48 21 (23) (23) 27 In a historic perspective, the reduction in process water consumption has been reduced significantly, especially for market pulp production. See appendix B for historical effluent generation data for the Swedish pulp and paper industry. The reported specific energy use for respective product is presented with the same production basis as for the above effluent volumes. The data has been gathered in the Energy Survey performed by Skogsindustrierna in cooperation with ÅF since the beginning of the 1970-ies [8]. 10 VÄRMEFORSK However, the complete energy consumption data for integrated kraftliner and newspaper is not easily available in issued reports. For these two qualities the below presented consumption data does not include energy used for pulp production and recovery of fibres for recycling. The consumption data is for kraftliner paper machine only. Table 2.2: Summary of energy consumption data for Swedish mills. Steam heat consumption year 2011 (ditto year 2000) (ditto year 1988) (ditto year 1973) GJ/ton - Power consumption year 2011 kWh/ton Market pulp 13.9 (15.2) (15.1) (16.6) Newspaper and magazine 5.1 (6.0) (6.7) (8.8) Kraftliner 4.5 (4.8) (4.4) (6.1) 796 776 483 (ditto year 2000) - (798) (711) (535) (ditto year 1988) - (837) (712) (472) (ditto year 1973) - (784) (786) (515) Information about the energy usage specifically related to cooling systems and/or secondary heat systems is difficult to find for Sweden or in any international literature. It is clear that the steam heat consumption has been reduced considerably since the beginning of the 1970-ies. During the same time period the water consumption has been reduced significantly due to a higher degree of process closure, see chapter 1.2 and Appendix A. Higher degree of closure generally increases the temperature in the process and reduces the need for steam heating (primary heat usage). Today’s lower water consumption for washing of pulp may to some extent explain the reduced primary heat consumption; also more energy efficient recovery of secondary heat leaving the process has contributed. The specific power consumption in the process has not been reduced since the beginning of the 1970-ies. This is most probably also explained by the increased closure of the pulp and paper mills. Even though the power consumption related to water consumption has gone down, the consumption has increased due to more extended effluent treatment and other additional process stages related to higher product quality demand. 11 VÄRMEFORSK 2.4 Energy and water consumption in the reference pulp mill The “Reference mills”, also in some context referred to as “Model mills”, were developed 2010 in a cooperation between ÅF and Innventia and are updates of the hypothetical reference mills developed in the FRAM and KAM project to reflect the technical changes that have occurred since 2005 [5] [6]. Four different types of pulp and paper mills were considered:  Bleached market kraft pulp mills – one softwood mill (pine), and two hardwood mills (birch and eucalyptus)  Integrated fine paper mill, with the pulp mill producing softwood and hardwood pulp in campaigns  Kraftliner mill  Magazine paper mill, bleached super calendered (SC) TMP In this project the focus has been on water and energy conservation possibilities for production of bleach market kraft softwood pulp, kraftliner and magazine paper. Therefore, the design of the water treatment and distribution system for the reference mill as well as the secondary heat system has been defined to a more detailed level within this project. A more detailed description of the reference mill can be found in Appendix C for bleached market pulp, Appendix D for kraftliner and Appendix E for magazine paper. 12 VÄRMEFORSK Key design data and process concept are summarized for the selected products. The more detailed water treatment and secondary heat design are also presented in Appendix C, D and E. Table 2.3: Energy and water consumption key data for selected reference mills. Market pulp7 Magazine paper 5.8 0.8 - Kraftliner Process water effluent m3/ton 22 13.7 ditto COD kg/ton <9 4 3 Cooling water m /ton effluent ditto heat load GJ/ton 7.6 1.2 2.1 Steam heat GJ/ton 8.9 3.3 9.1 consumption whereof paper mill 3.3 3.5 Power consumption kWh/ton 6008 2410 846 whereof paper mill 600 425 There is not cooling water effluent in the reference mills because concept includes recovery of all water to cooling towers. The difference between an average Swedish mill and the reference market pulp mill is large when it comes to water use, 22 m3/ADt for the reference mill and 77 m3/ADt for the average Swedish mill. Based on reported water consumption, we still consider that Swedish mills have process water consumptions and process effluent flows in line with international standards and BAT. The reference mills are theoretical greenfield mills using best available commercially proven technology. Each unit operation should be tested and key process data is conservative and should not exclude any major pulp mill equipment supplier. The use of wash presses in brown stock and bleach plant is keeping the use of fresh water down. Still it should be noted that the water consumption in many existing mills still is significantly higher than in the developed reference mills and one reason for this is the reference mill being a greenfield mill. For a mill that would like to decrease the water use there is thus a number of opportunities. The integration of cooling water and process water is one such option. Others are minimising the spills, use the condensate from black liquor evaporation more extensively and implement alkaline closure if not already implemented. 7 Softwood The specific power consumption of the reference mill has been updated based on new input from the latest commissioned market pulp mills. 8 13 VÄRMEFORSK 3 Process closure of the bleach market pulp mill This chapter discuss process concepts and measures for reducing the process water consumption of the “reference mill” for bleach market pulp described in Appendix C. After the screening the study has selected to further review further closure of the bleach plant. 3.1 Overview The pulp and paper industry is one of the major water consuming industries. A normal Swedish pulp mill consumes around 35 m3/ADt process water and discharge 16-18 kg COD/ADt to the recipient. The process water consumption can be reduced by process integration. It can be a combination of matching the process effluent of different contamination degree with a suitable water consumer, and internal water treatment before reuse. The reference mill model for bleach market pulp has been used to evaluate further closure of the bleach plant. This mill consumes 22 m3/ADt process water and the amount of COD from the bleach plant to external treatment is 17.8 kg/ADt. An increased degree of system closure can create problems with scale formation within the bleach plant and evaporation plants, high bleaching chemical consumption, corrosion and plugging problems in the recovery boiler and problems to control the Na/S balance of the mill. A simulation model of the reference mill was used to simulate the effect on further closure of the bleach plant. Among the analysed effects are; risk of precipitation of calcium oxalate, barium sulphate and calcium carbonate; increased chloride dioxide consumption due to increased COD in the bleaching stage; effluent volume and composition; need of chloride removal in recovery cycle; increased lime mud purge and make up in the mill. For many Swedish mills a lot of water savings could be done before the water use in the bleach plant is in the same order of magnitude as in the reference mill. Alkaline closure is not applied everywhere and the extensive use of condensate in the reference mill is also something Swedish mills could consider. 3.2 Further closure of the reference mill bleach plant 3.2.1 Reference mill In the reference mill the bleach plant is by far the largest process water consumer. Based on experience a relatively conservative approach regarding system closure has been adopted to ensure sustained trouble free operation with good economics in the reference mill. Fresh water is used for washing in EOP and in the drying section and white water is then used in the bleach plant for washing on the P stage wash press. A small part of the white water is discharged to the effluent treatment. 14 VÄRMEFORSK Wash presses are consequently used through the brown stock wash and the bleach plant. The brown stock washing is counter current without any bypass streams to the evaporation. The bleach plant is designed to release 14 t/ADt of effluent. The main part is acidic filtrate but the effluent contains also alkaline filtrate and a small white water stream. Alkaline closure is applied with recycling of EOP filtrate to the liquor cycle by the use of EOP filtrate on the last press in the brown stock wash before the bleach plant. This range includes a consumption of 5 t/ADt of fresh water in the bleach plant and 6 t/ADt in the drying machine. Both secondary A and B condensate is used in the bleach plant. When condensate B is used it has been purified by stripping to obtain the same COD content as in secondary condensate A, 200 mg/l. White water from the pulp dryer is used as wash liquor on the wash press after the Pstage. The filtrate from this wash press is then used as wash liquor on the D1 stage wash press. Fresh water is used as wash liquor on the EOP stage wash press and condensate is used as wash liquor on the D0-stage press together with a small amount of fresh water. The filtrate from the EOP wash press is then transferred as wash liquor to the 2nd wash press after the oxygen stage. The reference mill is described in more details in Appendix C. White water 1,8 Condensate A Hot water Hot water 0 4,6 1,1 4,0 White water 4,5 4,5 4,6 Condensate B EOP-stage D-stage Wash presses Hot water 0,5 Wash presses Bleached pulp storage tank D1-stage P-stage Wash presses Wash presses Wash presses To Brown stock washing 4,6 8,3 5,1 0,5 3/ Figure 2: Bleach plant filtrate use in the reference mill m ADt. Chemical water included in output streams. Secondary condensate B is stripped to the same COD content as secondary condensate A. 3.2.2 Decreased water use by increased filtrate use To decrease the water consumption in the bleach plant increased closure is necessary. The first thing to consider is to use D1 filtrate for washing on the wash press after D0. The condensate used for washing on the wash press after D0 could be used to replace the hot water on the EOP wash press together with a small amount of wash water. The effluent volume is significantly decreased from 14 to 9 ton/ADt. This was modelled in alternative A. 15 VÄRMEFORSK White water White water 1,8 0,5 Hot water Hot water 0 1,1 Condensate A 4,0 Condensate A 4,6 Hot water White water 0,0 4,5 4,5 4,6 Condensate B EOP-stage D-stage Wash presses D1-stage P-stage Wash presses Wash presses Wash presses Wash presses Bleached pulp storage tank To Brown stock washing 4,6 8,3 0,6 0 Figure 3: Flow strategy in bleach plant for alternative A. Chemical water included in output streams. Secondary condensate B is stripped to the same COD content as secondary condensate A. Another option of closure is to use counter current washing in bleach plant with alkaline and acidic filtrates and let the acidic filtrate go to effluent treatment and the alkaline filtrate to the brown stock washing. This alternative, B, is similar as alternative A but the condensate is then used on the wash press after D1 instead of at the wash press after EOP. The decrease in the filtrate volume to effluent is the same as for alternative A, 14 to 9 ton/ADt. White water White water 1,8 0,5 Hot water Condensate 0 Hot water Condensate A Hot water 1,1 Hot water 4,0 4,6 4,5 0 0,0 4,5 4,5 Condensate B 0,1 EOP-stage D-stage Wash presses White water Wash presses D1-stage Wash presses Bleached pulp storage tank P-stage Wash presses Wash presses To Brown stock washing 4,6 8,4 0,6 0 Figure 4: Flow strategy in bleach plant for alternative B. Chemical water included in output streams. Secondary condensate B is stripped to the same COD content as secondary condensate A The used amount of condensate and white water and effluent volume are exactly the same for alternative A and B but COD to effluent and carry over to the brown stock will be different. 16 VÄRMEFORSK 3.2.3 Results For both alternative A and B the chloride to the liquor cycle will increase. Alternative B gives higher chloride levels as more chloride is sent to the brown stock. To prevent problems in the recovery boiler a chloride removal process should be implemented. ESP dust leaching technique is becoming more and more common and is included in the standard equipment in the reference mill for eucalyptus but not for softwood. Introduction of the leaching process was simulated and presented as alternative A and B leach. In these cases the chloride level in the white liquor is kept constant by leaching part of the ESP dust. The sulfidity on the other hand is controlled by removal of unleached ESP dust. The summary of these two purges is presented in the table below. Table 3.1: Consequences of closure alternative A and B on the chemical recovery. Purged ESP dust Lime mud purge NaOH make up CaO make up Cl in white liquor Cl in Black liquor as fired K in Black liquor as fired Frac of total amount of ash to ash leaching Total amount of ash leached Total amount of bleach plant effluent Total COD in bleach plant effluent REF A 13.2 13.5 25.7 22.7 8.5 7.6 12.1 10.8 1.5 2.6 0.5 0.9 1.3 1.3 0,0 0.0 0.0 0.0 A leach 14.0 22.7 8.1 10.8 1.5 0.5 0.9 12.9 22.4 B 13.9 25.0 5.8 11.9 3.1 1.1 1.3 0.0 0.0 B leach 14.7 25.0 6.7 11.9 1.5 0.5 0.8 19.7 35.2 t/ADt 13.9 9.0 kg/ADt 17.8 16.8 9.0 16.8 9.0 16.2 9.0 16.2 kg/ADt kg/ADt kg/ADt kg/ADt g/l w% w% kg/ADt The results show that without ESP dust leaching the chloride concentration in the white liquor will increase from 1.5 to 2.5-3.1 g/l in the white liquor. With leaching of a reasonable amount of precipitator ash the concentration in white liquor (and also black liquor as fired) could be maintained at the same level as before. Potassium is also removed in the dust leaching and the concentration will decrease. Another option is to increase the dust removal without any precipitator dust leaching until the chloride levels are acceptable. This will increase cost for make up considerably but might be an option if the investment costs has to be kept down. NaOH make up is decreased due to more recovery of sodium from the bleach plant, more for alternative B than for alternative A compared to the reference case. The lime make up will be decreased, most in alternative A but also in alternative B. This is due to the fact that less fresh water is used in the process and in that way less NPE, especially Si, is added to the process. Fresh water contains relatively large amount of Si. Hence, the amount of Si to the recovery area is actually somewhat lower than in the reference case and therefore a lower lime mud purge is needed to obtain 90 % free CaO. 17 VÄRMEFORSK Figure 5: COD carry over and COD to effluent Both alternative A and B decrease the amount of COD to effluent. Even though the effluent stream from D1 is removed by using the filtrate for washing on D0, the decrease in COD to effluent is relatively small. The reason is that only a small amount of COD will be recycled to the recovery area. Major part of the COD in the D1 filtrate will end up in the D0 filtrate going to external treatment or in the carryover from the brownstock washing finally ending up in the D0 filtrate. The carry-over of COD to the bleaching stages will affect the consumption of bleaching chemicals. For alternative B the COD carry over to the D0 and EOP-stage is higher. The carry over to the pulp machine is a quality indicator and should not be increased. For alternative B it is decreased, main reason is the decreased carry over from the P-stage due to the condensate used on the press after D1 instead of P filtrate. The risk of precipitations should be compared to the reference case. In the figures below it can be seen that the reference mill has a risk for BaSO4 precipitation in the D0 tower and filtrate tank, but the amount is very low and the risk is minor. For CaCO3 precipitation there is a risk in EOP tower, EOP filtrate and P tower. CaCO3 is known to require a high degree of super saturation before any precipitation occurs and as the degree of super saturation is low it seems unlikely with precipitations in the P tower for the reference mill. 18 VÄRMEFORSK Figure 6: Simulated precipitations of BaSO4 and CaCO3. There is an increased risk of BaSO4 precipitations with alternative A or B in the D0 filtrate. For CaCO3 the increased risk is more obvious for alternative B than alternative A, especially in EOP filtrate where it more than doubled. No risk of calcium oxalate precipitations was indicated for the reference mill and alternative A and B in all positions in the bleach plant. Great care should be applied when interpreting and using simulated risk of precipitation. The data shall not be considered as absolute and true values. For instance it is known that the applied system of chemical equilibrium reactions overestimates the risk for scaling. This means that the model may predict that scales are formed, while they are not present in reality. Precipitation may also occur on the pulp. This could happen without any notice taken regarding pulp quality or other process related problems. However, modelled precipitations may serve as a guideline for points manual inspections of the mill during shut-downs. In the simulations only one wood composition was used. The content of non-process elements in the wood has a significant influence in the content in the process streams and the risk of formation of precipitates. A wood with a high content of NPEs will have a higher risk of forming precipitates Magnesium can be added to the bleach plant to prevent calcium oxalate precipitation. Magnesium and oxalate form a soluble complex and the consequence is decreased availability of oxalate and decreased risk of calcium oxalate precipitations [9]. In the reference mill talc is used in the brown stock and with the talc magnesium is added to the process. To verify that the low risk of precipitation is not caused by the use of talk, simulations were made without talk for alternative A. Still no calcium oxalate was formed. The conclusion from the simulation of the risk of precipitations is that alternative A is favourable. According to the simulations there is no reason to dissuade from increased closure with respect to increased risk of precipitations. However, simulations are always based on theoretical models and should not be seen as truths but more as guide lines. 19 VÄRMEFORSK The extensive use of condensate in the bleach plant assumes a stable and reliable quality of primarily secondary condensate A. Moving the position where the condensate is added towards the final product and remove the addition of water in the bleach plant could be risky, especially if the quality of the condensate is not stable. The stripping of condensate B to low COD content also has to be consistent but the position and amount used is the same for both alternative A and B as in the reference mill. The operability of the bleach plant and drying machine for alternative A and B is foreseen to be more challenging. Disturbances in the production rate from the digester to the drying machine cause imbalances in consumption of condensate, filtrate and white water. Generously sized buffers are required to maintain a high degree of closure also considering short term disturbances in the process. In the investment calculations this was considered with an assumed cost for buffering tanks of 5 MSEK. 3.2.4 Internal dilution before D0 An additional option to decrease the effluent from the bleach plant is to use filtrate from D0 to dilute the pulp to correct consistency prior to D0 tower. Another use of condensate and white water has to be considered for this alternative to be an option. It will not decrease the overall water use in the mill if the left-over streams could not be used to replace fresh water elsewhere in the mill. Figure 7: COD carryover and COD to effluent for reference mill, case A, case B and case A with internal dilution (A dil). Results from simulation of alternative A with internal dilution show a significant increase in the carryover to D0 tower. COD to effluent decreases but the risk of precipitations (calcium oxalate, calcium carbonate and barium sulphate) increases significantly. The chloride to recovery cycle is also increased compared to alternative A as there is a higher carryover to EOP. This is due to increased chlorine dioxide charge needed to compensate for increased carryover of COD and increased recirculation leading to build-up of chloride levels. Based on these results and the fact that there is no obvious use for the replaced condensate and white water, this process change is not considered as an interesting option for further closure of the bleach plant. 20 VÄRMEFORSK For Swedish mills the main benefit related to bleach plant closure is the reduction of COD in the effluent and the significant reduction in effluent volume. The reduced hot water consumption has some energy related value but the fresh water availability itself is seldom a technical problem. From marketing point of view the “water footprint” has is expected to have some value. Consequently, further closure as described in alternative A and B does not necessarily include extensive use of the evaporation plant condensate if pulp quality does not permit. In these specific cases the excess condensate is sent to sewer and hot water is used. 3.2.5 Additional possibilities. In all the closure alternatives above has the water to pulp machine and dilution factor in the bleach plant been the same. It is not recommended to use condensate close to the final product so the water used in the pulp machine should not be replaced by condensate. The risk of contamination is large if condensate is used and if the condensate quality for some reason drops it might ruin the quality of the final product resulting in loss in production and increased operating costs. On the other hand, if the dilution factor could be decreased, the result would be somewhat increased carryover in the final product which may be accepted for some qualities. A decreased dilution factor will save both fresh water and decrease the effluent volume. The decreased dilution factor could also be applied elsewhere in the bleach plant for decreased effluent volume without increased evaporation load. However, the consequences have to be evaluated carefully. For many Swedish mills a lot of water savings could be done before the water use in the bleach plant is in the same order of magnitude as in the reference mill. Alkaline closure is not applied everywhere and the extensive use of condensate in the reference mill is also something Swedish mills could consider. 3.3 Return acid filtrates from bleaching to liquor cycle The most extreme strategy would be to operate the bleach plant in a strict countercurrent mode regarding recycling of wash filtrates. The whole pulp washing sequence from digesters to pulp machine would then operate in countercurrent mode. Max water saving potential would be approx 5 m3/ADt as hot water provided that condensates and white water filtrate can be used elsewhere in the process. The water saving using strict counter current washing is the same as for increased closure alternative A and B described in 3.1.2 above while the effluent volume will decrease significantly. Mixing of acidic and alkaline filtrates should in general be avoided and several problems related to NPEs could be expected both in the bleach plant and in the recovery area. The NPEs dissolved from the pulp in acidic stages will eventually end up in the recovery cycle. Scaling in the evaporation is a risk that need to be considered. Some of 21 VÄRMEFORSK them will precipitate in the green liquor and with proper separation in the green liquor filtration end up in the green liquor dregs. If not they will accumulate in the lime cycle or end up in the white liquor depending on the solubility properties. Al, Si and P will precipitate in the lime cycle resulting in increased dead load and in order to not decrease the free CaO more lime need to be purged [10]. Operating and capital costs resulting from chemical consumptions, heating and cooling and the significant internal measures required for operating the mill at high availability are expected to be high. Energy consumption is expected to be significantly increased and the overall environmental effect is doubtful. This alternative is therefore not selected for further technical and economical evaluations in this study. 3.4 Separate evaporation and incineration of acid filtrates from bleaching Evaporation of bleach plant effluents and incineration of the solid residue can be a separate technology or combined with the strategy outlined in 3.3. The condensate produced could, after appropriate conditioning possibly including stripping and cooling, be recycled to raw water intake. However, the high risk for severe scaling of evaporator surfaces and the difficulty of handling evaporated solids are challenging issues. The heat required for evaporation decreases potential to sell power. The resulting requirement for cooling in surface condensers and recycled evaporator condensates would significantly increase the overall cooling demand in the mill. This alternative is therefore not selected for further technical and economical evaluations in this study. 3.5 Partial recovery of process effluent to water treatment plant or selected positions in the process Recovery of treated effluent to raw water intake or selected positions in the mill could reduce intake of fresh water. In fact, the Visy kraftliner in Australia and Canton market pulp mill in USA already operate by replacing part of the fresh raw water with recirculated treated effluent. These two mills are some of the most closed mills in the world and are described further in appendix A. The Visy paper kraftliner mill produced unbleached pulp and no bleaching chemicals are introduced to the process effluent. This makes the further treatment and recycling to the water treatment plant much easier compared to a bleach market pulp mill which is discussed in this chapter. When recycling treated effluent to raw water treatment the concentration of NPE need to be considered. There are many NPE such as Cl that cannot be removed in the external treatment and will accumulate in the process. 22 VÄRMEFORSK 3.5.1 Recovery to water treatment plant Recirculation of treated effluent to the water treatment plant could be feasible if the effluent has similar level of pollutants as the fresh water. However, very seldom the discharged effluent has the same degree (or lower) of pollutants as the fresh water. Another possible case is if the water treatment plant precipitation stage has a good capacity margin and could to some degree compensate for the higher degree of pollutants, although with an increased operational cost. As mention for the option above, the chlorides will not be removed in the external treatment and need to be considered before recycling the treated effluent to the process. Also other non-process elements will be returned. Note that there is no saving in operating cost related to recirculation of the effluent. The water treatment plant would act as a tertiary effluent treatment stage with similar operating cost and sludge disposal problem. The cost of further treatment of the process effluent with a third stage chemical precipitation to achieve desirable concentration before recirculation is analogous to typical high cost related to achieving low effluent emissions. Generally, further treatment and recirculation of effluent to the water treatment plant is foreseen to be relevant mainly for mills that are limited specifically on fresh water consumption or has a very poor fresh water quality. In Nordic conditions the fresh water availability and quality is normally not a problem, it is mainly the effluent volume and concentration of pollutants that is a concern from an environmental point of view. Any reduction of process water and energy consumption is judged to be limited in the Nordic countries. In the future water footprint can be more important and the situation can be changed. 3.5.2 Recovery directly to process One way of direct use of recycled effluent can for example be to replace fresh water to causticizing. Normally mills already use B-condensate in causticizing instead of fresh water. This is also the reference mill design. Recovering effluent to the causticising would therefore not save fresh water and the befit would be limited to lower COD load on the effluent treatment. Depending on the ionic concentration in the effluent, additional systems such as e.g. ash leaching could be required. Since this principal technology needs additional specifications requiring other technologies in order to fulfil the quality-specification for the intended water use, it is considered to be out of scope of this study and therefore not selected for further technical and economical evaluation. 3.6 Membrane filtration of effluents and internal reuse Membrane separation technologies are discussed in Appendix A. Lately a number of studies have been investing how process flows in pulp and paper mills can be treated by 23 VÄRMEFORSK combinations of membrane filtration, flocculation and anaerobic stages in order to achieve more efficient energy and resource utilization with low discharges [11] [12]. The possibility of treating an effluent flow in membrane filters and recycle the permeate (low COD concentration) to the process and further treatment of the concentrate (high COD concentration) in anaerobic treatment and aerobic treatment has been studied year 2011 [11]. The process integration reduces the water consumption, as well as decreasing the sludge disposal because of biogas generation from the concentrate. Also a slightly modified concept where the concentrate is returned to the process and incinerated in a boiler has been reviewed year 2013 [12]. The permeate is processed in an anaerobic treatment and a following aerobic treatment. The technology has been studied for a number of pulp and paper products. One product analysed is bleached softwood kraft pulp in a mill producing 500 000 ADt per year. The alkali effluent and spills from the fiber line are sent to a membrane filtration unit. The report does not clearly present how the sludge is disposed. Disposal in a recovery boiler is foreseen to increase NPE accumulation, and disposal of sludge in power boiler is generally challenging. In the 2011 study the permeate is reused in the fiber line alkali bleach stage for substituting fresh water [11]. The concentrate is sent to anaerobic and aerobic treatment. The anaerobic treatment do also take care of the acid bleach plant effluent, evaporation plant condensate, wood handling effluent as well as the sludge produced in the anaerobic treatment. The estimated investment cost within the study is 152 MSEK yielding a net saving of 11 MSEK per year. The specific water consumption and corresponding effluent reduction was calculated to 12.6 m3/ADt relative to the base case effluent flow of 34 m3/ADt. In the 2013 study a “reversed” process integration was considered where the permeate is sent to the anaerobic treatment, and the concentrate sent to the evaporation plant before incineration in the biomass boiler or recovery boiler. The estimated investment cost within the study is 177 MSEK yielding a net saving of up to 12 MSEK per year. The specific water consumption was not reduced but the TOC load on the recipient was reduced significantly. Membranes can separate high molecular mass components from a stream but the ionic concentrations are unaffected. In line with reasoning under chapter 3.5, this technology requires evaluation for specific cases, possibly in combination with other technologies, and is not selected for further technical and economical evaluations in this report. 3.7 Alternative usage of evaporation condensates The evaporation condensates leave the different evaporator bodies with varying degree of contamination. The evaporation condensates are often grouped into three qualities: Acondensate, B-condensate and foul condensate. The A-condensate is very clean and is normally good enough for using in the oxygen delignification or possible the first part of 24 VÄRMEFORSK the bleach plant. The B-condensate includes some COD and traces of smell and is used in the causticising area where the process requirement is lower. The foul condensate, also referred to as C-condensate, is unclean and is therefore treaded in a dedicated stripper. The stripper removes the volatile substances and COD. The “stripped condensate” is comparable with A- condensate and can be used in the process, possible oxygen delignification and bleach plant if the end product quality accepts. The stripped condensate is often referred to as A- condensate since they have comparable quality. The B-condensate generation is typically adjusted to match the quantity that can be utilized in the causticising area in miscellaneous positions. B-condensate can be used in all the causticising positions except the lime mud filter cake washing which is extraordinary sensitive for residual odour escaping with the flue gas in the lime mud dryer. Therefore, hot water is used for lime mud washing but this consumption can be minimized by two stages washing where the clean hot water is limited to the final stage. The reference mill used B-condensate for above recommended positions, i.e. all except for lime mud washing. The A-condensate and “stripped condensate” often exceeds the amount that can be accepted for washing in the oxygen delignification and bleach plant. Many Swedish mills already use alkaline filtrate for washing in the oxygen delignification to reduce the COD load on the effluent treatment. Consequently, many times the evaporation condensate cannot be used in this “odour safe” position. The reference mill design assumed utilization of all A-condensate in the bleach plant. If usage in bleach plant is not accepted, the excess condensate is sent to sewer. The temperature of the stripped condensates is normally high and this heat is recovered before it exits the mill via the effluent treatment plant. Extensive usage of evaporation condensates in the bleach plant is rather uncommon in the Swedish pulp mills, and therefore many mills dump from this excess of clean evaporation condensate to the effluent treatment. Alternative use of this excess condensate is interesting because it would reduce the fresh water consumption and reduce the effluent volume (even though the reduction in COD is insignificant). Extensive usage of produced A-condensate and stripped B-condensate is discussed in chapter 3.2. There are references of mills using the excess evaporation condensates in the wood preparation. The condensate quality must be high enough to avoid significant odour problems. The wood preparation is generally a good water reuse “sink” that accepts lower quality water. Therefore this position may already be occupied by “low quality” water. The other alternative positions are represented by the bleach plant washing or drying machine hot water box and showers. As noted above the main restriction is related to product quality and the risk of contamination through odour and taste. 25 VÄRMEFORSK The evaporation condensate can be cleaned to very high degree if the stripper is properly designed with high capacity. Even though all the volatile COD related to smell and taste are removed it must be carefully studied if any other compounds such as phenols remains or ions from carry over droplets in the evaporator. The main challenge and risk of extensive utilization of evaporation condensates close to the end product is related to evaporation plant disturbances that could cause temporary contamination of the condensates which is not detected. It is questionable if reuse of Acondensate or stripped condensate can be accepted for washing in the final bleach stages for products such as liquid board. The modern systems can be designed to dump the condensates leaving the evaporators at as low conductivity as 300-500 uS/cm. There are also mills that use 4000 uS/cm as limit for dumping evaporation condensate. The principal idea of further treating and using evaporation condensates is interesting but requires more specific case studies in order for relevant technical and economical evaluation why it is not included in this study. 3.8 Reduced consumption and recovery of sealing water For a new market pulp mill the sealing water consumption can be about 1 m3/ADt. For existing older mills this quantity is expected to be higher because of generally larger number of pumps with older technology. Sealing water is typically produced in the raw water treatment typically through pressurized filters and stored in a separated tank. The sealing water is distributed mill wide at a common pressure level and if higher pressure is required there are booster pumps located in the different areas. There is also a concept where the sealing water is produced, stored and distributed locally in each area, but this is rather uncommon. The consumed sealing water exits via the low solids sewer. The sealing water can be alternatively collected via concrete channels in the floor and led to a basin in each area. It is pump back to the water treatment plant. Asia Pulp & Paper (APP) Hainan Jinhai is a single line pulp mill fine paper machine which has implemented sealing water recovery. The annual production capacity of 1 million tonnes of pulp and 900,000 tonnes of fine paper. The daily sealing water saving is reported as 3500 m3 corresponding to around 1.2 m3/ADt pulp [13]. A number of mills in South Africa, Australia and Brazil have implemented systems for recovering sealing water. Common for all these mills that have implemented recovery of sealing water is that there is a firm environmental permit limiting the effluent emissions or water consumption. 26 VÄRMEFORSK Another approach for reducing the sealing water consumption is to install double mechanical seals in pumps. This measure becomes more important to implement when the mill aims against very low water consumption and effluent flow. It is not foreseen to be a measure in Swedish mills because it cannot be justified environmentally or economically with the high water availability in Nordic conditions. Furthermore, the sealing water effluent itself does not include any contaminants such as COD. 3.9 Recovery of purge from cooling towers Pulp and paper mills designed with a closed cooling water system (not semi-open as the model mill) needs a purge from the cooling water circuit. The evaporation in the cooling tower concentrates salts and organics present in the cooling water, to keep these on a constant and acceptable level some of the cooling water is purged. The purge flow is determined by the mill water quality and level of pollutants that is accepted in the system. The reference mill is designed with a semi-open system and does not suffer accumulation of pollutants in the cooling water system. For mills not designed with a semi-open system it is although interesting to review the possibility of reusing the purge from the cooling towers. It should although be noted that the reused purge contains NPE:s that can cause problems in liquor cycle of a kraft market pulp mill. Visy paper has a very low water consumption and effluent flow but has selected to not reuse the purge from the cooling towers. Water used to ‘backwash’ the cooling tower filters is in Asia Pulp & Paper (APP) Hainan Jinhai now reused in the recausticizing process to produce white liquor – generating a daily saving of 2,300m3 [13]. It is questionable if this measure is relevant to implement if the mill is already very closed and suffers from problems with NPE:s in the liquor cycle. 3.10 Collection of steam condensate from pipe racks Collection of steam condensate from pipe rack is rather uncommon. The condensate losses from traps are generally very low, possible in the order of 0.2 m3/ADt for a market pulp mill. The condensate trap losses have although increase for new mills which is designed for wet LP steam extracted from the high efficiency turbines. This measure is planned to be implemented in one of the new eucalyptus pulp mills in South America. The measure is manly justified by the improved visual impression with recovery of the condensate from the pipe racks. Cold steam condensate is used to 27 VÄRMEFORSK quench the hot steam condensate from the traps and the mix is pumped back to the boiler house. Asia Pulp & Paper (APP) Hainan Jinhai has announced that they have installed water condensing on the facility’s pipe racks. The steam condensate is collected and reused in the facility’s boilers generating steam for the steam turbines and saving 600 m3 daily equal to around 0.2 m3/ADt [13]. 3.11 Returning cooling water from oil coolers The cooling water consumption in oil coolers is in the range of 0.5-1.0 m3/ADt in modern market pulp mills. Typical oil coolers are located in the turbine area and air compressors for production of instrument and mill air. This water is normally not returned to the cooling water circuit because of the risk of contamination. The flow instead discarded and is blended with the treated process effluent before escaping the mill. However, with a limited investment in oil separators, it is possible to guarantee that clean cooling water is returned to the cooling water circuit. This measure is relevant to consider if the mill needs to reduce the raw water consumption further or limit the effluent volume. Oil separators have been installed in mills with an unusually tight environmental permit. Another alternative is to return the cooling water from the oil coolers to process area such as wood preparation where any possible oil contamination would not cause substantial disturbances to the process. This concept has been applied in at least one of the new green field market pulp mills in South America. 28 VÄRMEFORSK 4 Process closure of the kraftliner pulp mill This chapter discuss process concepts and measures for reducing the process water consumption of the “reference mill” for producing kraftliner pulp described in Appendix D. After the screening the study has selected to not further review any of the discussed alternatives. 4.1 Overview The kraftliner process water system is rather similar to the market pulp mill. The main difference is related to the water usage on the paper machine and RCF-plant, as well as the non-existing bleaching plant. For further info see Appendix D. No bleaching (or possible partial) of the produced chemical pulp allows counter current washing at alkaline conditions with recovery of the filtrate into the liquor cycle. This is difficult to accomplish in the market pulp mill where chlorine containing chemicals often is used in the bleaching process. Also the pH is lowered to acid conditions there many non-process elements (NPE:s) are dissolved from the pulp and will be troublesome if the filtrate is recycled back to the brownstock washing. On the other and, the white water system and water chemistry in the kraftliner paper machines is more complicated to manage compared to a market pulp dryer. It is not possible to use the white water for counter current washing in the pulp mill without adapting the chemical usage on the paper machine. If the kraftliner paper machine is integrated with RFC plant, the white water can be used in the RFC plant before sent to the effluent. 4.2 Measures for the market pulp mill that are valid for kraftliner Several measures discussed for the market pulp mill in chapter 0 are also valid for the kraftliner mill. A summary table presented in chapter 7.1 presents a list of measures that could be implemented in both mill types. 4.3 Complete evaporator condensate reuse The process water consumption can be reduced by using all evaporation condensate in the process. About 2.2 m3/ADt evaporation condensate of totally 6.8 m3/ADt is utilized in the causticising area similar as done in the reference market pulp mill. The remaining 4.6 m3/ADt can be used for pulp washing, this eliminates possible hot water consumption for washing in the pulp mill. Analogous to the market pulp mill, the reliability of an odour free A-condensate quality and the product quality needs to be considered. 29 VÄRMEFORSK 4.4 Using paper machine white water for pulp washing If all evaporation condensate is reused within the pulp mill (full recovery possible for certain kraftliner product qualities), next potential step for extreme closure of the mill requires reuse of kraftliner paper machine white water for pulp washing. The reuse of white water in the pulp washing makes the clean evaporation condensate available for usage on the paper machine or other mill areas. Also the white water effluent from paper machine is reduced. Integrating the paper machine white water with the pulp mill washing has only been identified in Visy paper kraftliner mill. The closure of Visy paper is extreme and is described further in appendix A.5. 4.5 Membrane filtration of effluents and internal reuse Alternatives for reducing effluent flow and fresh water consumption has been studied for a number of pulp and paper products [12] [11]. This has however not been studied specifically for kraftliner production. The general concept and conclusions are described in chapter 3.6. It is expected that the kraftliner mill do not have any specific advantages that would result in a significantly better result compared to the products studied so far. 30 VÄRMEFORSK 5 Process closure of the magazine paper mill The reference magazine paper mill has a fresh water consumption of 6.5 m3 per ton paper and an effluent flow of 5.8 m3 per ton paper, and is therefore relatively closed and uses the best available technology, for further description see Appendix E. The possibility of reducing the water consumption further is thereby fairly limited. A few options of reducing the process water consumption further have been identified: 1. Substitute fresh water used in chemical preparation with super clarified filtrate from paper machine. 2. Reduced consumption and recovery of sealing water as discussed in chapter 3.8. 3. Recover cooling water from hydraulic oil coolers to process through oil separators 3.11. 4. Substitute fresh water used on the paper machine high pressure showers with super clarified filtrate. 5. Impregnation of chips for uniform moisture content before refining with white water instead of fresh water. Reducing the water consumption in the paper mill may reduce the operation cost and certainly the environmental foot print. However, higher closure of the mill increases the built-up of non-process elements (NPEs). The runnability and product quality of the paper machine is decreased because of problems with water chemistry and increased microbiological activity. The main part of the fresh water is used on the paper machine to reduce the above discussed problems at high degree of closure. The resulting surplus of white water at the paper machine is used counter currently in the pulping process before exiting the system via the effluent treatment. Consequently, minimizing the fresh water consumption on the paper machine also reduces the surplus of white water that can be used for washing the pulp counter currently in the TMP mill resulting in higher carryover of COD to the paper mill. Installing a system for recovering the sealing water and cooling water from hydraulic units can be done without affecting the built-up of NPEs in the process. Total potential is estimated to be in the order of 0.5-1 m3/ton paper. If implementing the other measures it is necessary to introduce a kidney for removing NPE:s and “stickies” to achieve a higher degree of closure of the reference mill. A number of internal and external treatment technologies are available for producing clean water that can be recirculated to the process:    Membrane filtration of super clarified filtrate from disk filters and reused on the paper machine. Membrane filtration of water from the disk filters after refining and screening. Recirculation of treated effluent to process primarily for usage within the pulp mill for washing and dilution instead of white water from the paper mill. 31 VÄRMEFORSK  Recirculation of treated effluent to water treatment plant for further polishing before reuse in process. Evaporation of effluent with surplus steam or in a mechanical vapour recompression (MVR) evaporation system. Combustion of strong liquor in recovery furnace or power boiler. The above separation technologies are required for reducing the reference mill fresh water consumption to a minimum without compromising the runnability and product quality of the paper machine. 5.1 Membrane filtration of effluents and internal reuse Increased closure of the mill by membrane filtration is discussed for bleached market pulp in chapter 3.6. The application of membrane filtration for integrated mechanical pulp mills has also been studied [11] [12]. The basis for the study was production of bleached TMP pulp with integrated paper production of 300 000 t/year . In the TMP mill the wood chip wash water is sent together with excess white water to a membrane filtration unit. Excess of white water in the paper mill is also sent to the membrane filtration unit. The permeate (low COD concentration) is recycled to the process and the concentrate (high COD concentration) is sent to an anaerobic treatment and aerobic treatment [11]. Also a second slightly modified concept where the concentrate is returned to the process and incinerated in a boiler has been reviewed [12].The permeate is processed in an anaerobic treatment and a following aerobic treatment. The investment cost for the first alternative was estimated to 122 MSEK and the net operation cost reduction was 11 MSEK per year [11]. The investment cost for the second alternative was estimated to 106 MSEK and the net operation cost reduction was 7 years. [12] The specific water consumption and corresponding effluent reduction was calculated to 11.2 m3/ton paper relative to the base case effluent flow of 16 m3/ton paper. The second alternative did not reduce the water consumption and effluent flow but reduced TOC load on the recipient. 5.2 Recovery of effluent to water treatment or selected positions in the process There are two known paper mills in Sweden which recirculate process effluent to the process or water treatment plant. One of these is Munkedal mill (see Appendix C). Chapter 3.5 discuss the possibly of reusing treated effluent in chemical pulp mills. 32 VÄRMEFORSK 5.3 Evaporation of effluent from TMP plant The reference magazine paper mill has a surplus of 1.45 GJ/ton paper steam heat could be used for evaporation of effluent. This surplus would be enough excess heat to evaporate the entire effluent of 4-5 m3/ton paper effluent to a concentration of 30-40% dryness. There are some references of mills concentrating parts of the effluent through evaporation. The evaporation condensate could be reused in the pulp washing. The produced strong liquor can be incinerated in a dedicated furnace, or alternatively sent to any other nearby boiler capable of firing the liquor. As an example, since 1992 the CTMP mill Millar Western Meadow Lake’s located in Canada achieves a freshwater intake rate of 2.5 m3/ADt and has eliminated the process effluent discharge by concentrating the effluent in an evaporation plant. The produced liquor is incinerated in a dedicated furnace and the smelt is sent to the landfill. The produced condensate is reused in the process. Another example is Fors CTMP and board mill in Sweden which concentrates a part of the process effluent in an evaporation plant. The produced concentrated liquor is incinerated in a nearby chemical pulp mill [7]. The investment cost for an evaporation plant is fairly high and the operation cost related to the steam consumption can be significant. However, the reference mill has a surplus of steam that is not used. 33 VÄRMEFORSK 6 Optimized secondary heat and cooling system 6.1 Introduction Primary heat is consumed in the production process and the heat that can be recovered is called secondary heat. The secondary heat system and cooling systems starts where the primary heat is transferred to the process. The primary heat, secondary heat and cooling systems are all therefore interconnected. Utilization of secondary heat can to some extent replace primary heat and thus reduce energy consumption. Secondary heat can be warm media such as condensates from flashing liquors from digester and evaporation areas that directly can be used. Secondary heat must often be transferred to another media to be usable, e.g. through heating of water in a heat exchanger. The hot water system is normally the most important secondary heat system of the mill. The higher the temperature of the secondary heat, the easier it is to find a useful application and the more valuable it is. A properly designed hot water and cooling water system is one of the basic requirements for an energy efficient pulp mill. The economic value of secondary heat is usually assigned the value of saving primary heat. The true value can be anything in-between 0 and 1 in relation to the primary heat value. 6.1.1 Objectives of secondary heat and cooling water systems The cooling water system is integrated with the secondary heat system. Principally there are two objectives when heat exchanging two media: 1. Satisfy the process cooling demand to ensure operation and high plant availability. 2. Recover high value heat for utilization within the process or by external clients. The former normally has the highest priority because it may be directly related to the production. If the cooling demand in key position is not fulfilled the plant production capacity may be reduced, or the process must be stopped. The latter implies the importance of continuous work to reduce the primary heat consumption and the operating cost of the plant. The importance of efficient secondary heat recovery has grown with rising prices of primary heat (biomass fuel, fuel oil, electric power etc.). Lower energy consumption per produced unit of paper and pulp can also reduce the overall environmental impact. 34 VÄRMEFORSK In some cases complex cascade interconnections between heat sources contradicts ensuring a reliable cooling system and stable operation. A successful design requires that both the cooling demand and variations in the process are taken into account, combined with an energy efficient heat recovery. 6.1.2 Principles of efficient heat recovery The general principle is to cool hot streams with hot media and colder streams with colder media as far as possible minimizing the temperature difference (ΔT) between the streams. This is normally referred to as pinch analysis of the process. Specific pinch analysis studies have been performed for a number of pulp and paper products but are not commented in detailed in this study [14]. Pinch analysis is a good tool for improving the understanding of thermodynamics of energy systems, and is a method for defining the highest possible degree of integration in the process in order to minimize primary energy usage. Pinch studies have been performed within the industry; however practical aspects such as process variation and distance in-between heat sources and heat sinks can limit the applicability of the pinch analysis. Practical limitations related to the plant layout and reliable cooling capacity must to be addressed. One of the most essential outputs of a pinch analysis is the grand composite curve which shows how all process streams are heated and cooled within the system. From the grand composite curve it is possible to identify the “pinch temperature” for the system. The pinch temperature separates the system in two parts: above the pinch temperature there is a deficit of heat that must be compensated by transferring primary heat to the system, below the pinch temperature there is a surplus of heat that needs to be cooled away. 35 VÄRMEFORSK Figure 8: Iggesund mill composite curve published 2011 after performed pinch analysis [15] The blue line show all heated process flow with increasing temperature (used as cooling media).The red line show all process flow that are cooled with decreasing temperature (used as heating media) The Iggesund mill pinch temperature was calculated to be around 65-75°C [15]. A number of pinch analyses have also been published for Värö mill [16] [17]. A more general approach to optimizing the warm and hot water system by pinch analysis was preformed within the FRAM program [17] [18]. The objective was to maximize the amount of excess heat by moving the cold composite curve towards the hot curve in the composite curve diagram until a minimum acceptable ΔT was reached for the system. In 2007 a survey was done to estimate the “temperature requirement” for upgraded heat, analogous to calculating the “pinch temperature”. The survey was done for a number of Swedish pulp and paper mills [19]. The result is briefly discussed in Appendix A. 6.1.3 Definition of coolers and heat recovery units A heat exchanger can work as a “cooler” or “heat recovery unit”. This report distinguishes heat exchangers used only for cooling purposes, and heat exchangers working as combined coolers and heat recovery units. The principal difference is that heat from heat exchangers used as “coolers” is not recovered to the process; it leaves the system. 36 VÄRMEFORSK Heat from heat exchangers working as “heat recovery units” does not leave the system; it is at least partially recovered to the process. In this case the heat has an economical value. 6.2 Design to improve secondary heat system in mills 6.2.1 Methodology The first step is to determine the temperature level at which the mill suffers from a secondary heat deficit and where this deficit is turned to a surplus. Principally this temperature level is analogous to the pinch temperature definition described briefly above in chapter 6.1.2. A good indicator of the efficiency of a secondary heat system is the temperature of the water that is used in the bleach plant or equivalent big consumer. A high and even temperature indicates a well designed and well operated secondary heat system whereas a low and/or varying temperature indicates that there is room for improvement. In practice a deficit of secondary heat is indicated by live steam (primary heat) usage for hot water heating, or in the process as a consequence of too low hot water temperature. The following table describes the most common attributes of well functioning systems with a surplus of secondary heat compared to a poorly designed system with a deficit of secondary heat. 37 VÄRMEFORSK Table 6.1: Common attributes of for hot water surplus or deficit in mills. Situation 1: Deficit of secondary heat (poor design) The hot water system is split into at least two, preferably more systems at different temperature levels No cascade interconnections Continuous usage of live steam in hot water tank to be able to maintain temperature set point. Overflow of hot water tank is very uncommon. Low hot water temperature fiber line, paper machine and other mill areas (<75°C). High steam consumption in the bleach plant due to low temperature wash water. Steam consumption for heating white water in drying machine/paper machine. Continuous usage of live to heat evaporation condensate or fresh water Steam is used in a large extent to heat the district heating return. No or limited preheating of boiler makeup water. Situation 2: Surplus of secondary heat (well functioning system) Single temperature level for produced hot water The system is arranged in cascade so that low value heat sources heat the coldest water, while heat sources at a higher temperature level increase the temperature level increase the temperature in the following stage No steam is used for water heating in fiber line, caustizising or drying/paper machines There is frequently or continuous overflow of hot water from buffer tanks High hot water temperature in fiber line (>75°C) Low steam consumption in bleach plant by usage of up to 90-95°C wash water if possible. No steam consumption for white water heating at drying machine/paper machine No steam consumption for local heating of fresh water or evaporation condensate within causticizing Steam usage for district heating is uncommon, only for very cold winter days. Preheating of make-up water to around 75°C. When a poorly designed system with a secondary heat deficit has been identified alternatives for increasing hot water generation should be explored. This situation can generally be divided into two sub-cases:   The mill hot water temperature is low (<75°C) but there is an excess of hot water at this level in terms of volumetric flow (overflow in hot water tank). The focus is on producing hot water with higher temperature but not necessarily larger volumes (additional MWs). The mill hot water temperature is acceptable (>75°C) but there is a shortage of hot water at this level in terms of volumetric flow. There is seldom an overflow in the hot water tank, and make-up of cooled water that is heated with steam is required. The focus is on producing a larger volume of hot 38 VÄRMEFORSK water (additional MWs), not necessarily higher temperature. With a well functioning system with a secondary heat surplus alternative utilization of the excess secondary heat should be explored. Alternatively, the surplus of secondary heat allows further reduction of the primary heat consumption which as a consequence reduces the secondary heat generation. Typical measures to improve the secondary heat generation are discussed in 6.2.2. Alternatives for reducing the primary heat consumption by utilizing excess secondary heat are discussed in 6.2.3. 6.2.2 Improving heat secondary heat recovery and minimize consumption of A secondary heat system can be designed in many ways depending on the circumstances in the individual mill. With rising energy costs there is an increased importance of secondary heat to reduce primary heat consumption. Some measures for improving secondary heat recovery are discussed in the following chapter. 6.2.2.1 Review sizing of heat exchangers One of the most obvious and most frequently implemented measures is to increase heat exchanger sizing. If the same cooling media is utilized (say warm water 45°C) and cooling duty (MW transferred) is fixed, a larger heat exchanger would allow production of hot water with a higher temperature but with a lower flow. Alternatively, a larger heat exchanger would allow recovery of more heat (higher duty) while maintaining inlet and outlet temperatures on the cold side. The heat exchanger sizing is particularly interesting to review if the mill hot water temperature is low and there is still an excess of hot water in terms of volumetric flow. More generous sizing can then increase the hot water temperature and reduce or eliminate the overflow in the hot water tank. The higher hot water temperature reduces the need of steam. Undersized heat exchangers from a heat recovery point of view are fairly common since initial design conditions are no longer representative of actual operating conditions after continuous capacity creep and rebuilds. These “undersized” heat exchangers do still have enough capacity to fulfil the most important “cooling” duty to keep the process going, although resulting in high flow/low temperature of outgoing hot water. In a chemical pulp mill the weak black liquor cooler and primary condenser in the cooking area are the hot water producers that have the highest potential for improvement. The green liquor cooler and dissolving tank scrubber should also be subject of careful review. 39 VÄRMEFORSK On a paper machine the sizing of the blow-through condenser is important. The blowthrough condenser has the potential to reach a relative high warm water temperature. Warm water can also be produced in the dryer air exhaust heat recovery. For new installations it is important to review the sizing of heat exchangers and make sure that guaranteed performance is based on reasonable heat transfer coefficients. 40 VÄRMEFORSK Below is a principal example of how heat exchanger sizing affects the result if the same cooling source is utilized (say warm water 60°C) and cooling duty is fixed (1 MW transferred). The higher hot water temperature in the case of the larger heat exchanger could enable reduction of steam heating in different mill areas, or possible improve the conditions for utilizing excess heat for alternative purposes (district heating, biomass drying). Smaller heat exchanger Larger heat exchanger Figure 9: Principal description of heat exchanger sizing and performance for a given cooling duty. 6.2.2.2 Cascade interconnection of heat exchanger recovery units The second most common measure to improve existing systems and optimize new installations is the interconnection of heat recovery units. Generally, feeding high temperature heat recovery units with cooling water (5-25°C) should be avoided. These hot water producers should be feed with at least warm water (45°C) or preferably semi-hot water (65°C) to maximize the hot water production (85°C). However, this requires more heat exchanger area but it can normally be justified if there is a shortage of hot water in the mill. When interconnecting heat exchangers in cascade it is important to address the process variations and how these could affect the cooling and heat recovery. If possible a buffer tank in-between the heat recovery units would compensate for the process variations. 41 VÄRMEFORSK When interconnecting heat recovery units, the layout also becomes a more essential aspect because the piping costs can become high with long distances. A parallel configuration is less complicated from this point of view since there normally exists a distribution system for some typical temperature levels (warm water 45°C, hot water 80°C etc.). Figure 10: Principle description of parallel configuration of heat recovery units (to the left side) and a cascade configuration (to the right side). 42 VÄRMEFORSK 6.2.2.3 Monitoring heat exchanger performance Before examining possibilities of increasing heat exchanger capacity it is relevant to review the performance of the existing heat exchanger. This can be done by studying the heat transfer rate in the recovery unit and comparing to design and typical performance depending on media and heat exchanger type. The largest and most important heat recovery units should be monitored continuously and automatic cleaning performed when the U-values (overall heat transfer coefficient) decrease below a given set point. Continuous monitoring of heat exchanger performance has been implemented in Stora Enso Nymölla pulp mill. The mill has good experience of the system and reports significant improvements in the performance for the heat exchangers where it has been implemented [20]. 6.2.2.4 Monitoring secondary heat system performance Flow and temperature are the key parameters in the warm and hot water system that should be monitored continuously through flow and temperature measurements. Monthly follow-up of the system performance is recommended including evaluation of improvement options. Personnel on all levels should be well trained regarding both the proper function as well as the possible malfunction of the system 6.2.2.5 Preheating of fresh water Preheating a major portion of the cold water to 20°C in the winter eliminates seasonal variations in the water systems. The required heat can be partially extracted from cooling of effluent or recirculation of cooling water to the raw water treatment. 21% of the Swedish and Norwegian mills preheat the incoming fresh water (see further Appendix C). Systems designed with cooling towers are normally operated with a fixed distribution temperature on the cooling water circuit. In the winter cold make-up water from the raw water treatment feed to the cooling water distribution pumps for evaporation plant or to a possible condensing turbine to improve the cooling capacity in these specific processes. On paper machines cold make-up water is often used as cooling water/sealing water for vacuum pumps on the wet-end of the machine. In this position, the make-up water is preheated and at the same time the vacuum is improved and the power consumption is reduced. The warm water leaving the vacuum pump may be contaminated by fibres and should only be reused positions accepts lower water quality. 43 VÄRMEFORSK 6.2.2.6 Hot water temperature and size of hot water tank The hot water tank temperature should be kept at as high and constant as possible. For new pulp mills a typical design is to reach 85-90°C without need of any primary heat. A high hot water temperature enables lower steam heating of pulp in the bleaching stages. It also enables improved utilization of secondary heat when possible surplus hot water is used for boiler make-up water preheating, combustion air heating etc. The main limitation to keeping a high and constant hot water temperature is related to the system design and performance of the heat exchangers used for hot water production. The volume of the hot water tank is important for balancing the variations in production and consumption of process water. The hot water storage for new mills is normally 1000 m3. The volume of the hot water tank in existing mills can be very different from mill to mill. Also the water quality for the hot water system is important when operating at high temperature levels. Poor quality water causes precipitation and reduces the heat transfer capacity in heat exchangers. As for boilers, evaporators and steam and condensate systems, the risk for these problems increases with the temperature. 6.2.2.7 Storage of energy in hot water accumulator District heating systems often include a hot water accumulator to reduce the peak loads and the resulting consumption of fuel oil in the boilers. Normally pulp and paper mills do not experience the same degree of long term variation in the process heat consumption as district heating plants do due to weather and the daily changes in tap water consumption. The mill hot water tank (i.e. secondary heat accumulator) is normally sized for short term unbalances in the process. It is difficult to justify a larger hot water storage because mill availability is normally high and the production is stable. Furthermore, the mill hot water consumption is generally not affected by fluctuations in outdoor temperature. Hot water accumulators integrated in district heating plants can be up to 50 000 m3with significantly higher heat storage capacity. These large sized heat storages can economically be justified by reducing the peak fuel consumption which often is expensive fuel oil. Also an accumulator can reduce the minimum required boiler heat generation capacity by taking away peaks. If the pulp or paper mill is integrated with the district heating net it could be relevant to increase the secondary heat storage capacity. Such a heat accumulator could be integrated with the pulp mill secondary heat system. Heat could be charged either from the evaporation plant, flue gas cooler, weak black liquor cooler or any other high temperature secondary heat source. It can be unloaded against the district heating neat as well as the pulp and paper mill. The benefit of having a heat storage must be evaluated against the heat losses from the storage tank and the rather high investment cost. 44 VÄRMEFORSK In many cases it may be advantageous to have a hot water accumulator in a pulp and paper mill that is integrated with a district heating net. Especially if the mill is the main supplier of district heating, the power boiler capacity is limited, and there is not an existing accumulator in the system. 6.2.2.8 Increased hot water generation from weak black liquor If there is a hot water deficit it may also be relevant to review the possibility of cooling the weak black liquor to a lower temperature. Measures 6.2.2.1 and 6.2.2.2 are related to how the water side of, for example a WBL cooler, is optimized with the same energy (MW) transferred (same set point on WBL outgoing temperature. This measure discusses if it can be interesting to reduce the WBL temperature set point to increase the energy (MW) that can be transferred to the hot water side. Reducing the WBL temperature is interesting if there is a hot water deficit and live steam is consumed for hot water production or as a consequence of low hot water temperature. It is important to note that a lower WBL temperature would also affect the evaporation plant steam consumption in a negative way. However, the evaporation plant increase in LP steam consumption is always lower than the equivalent LP saving for hot water production. The reason is that WBL entering the evaporation plant is heated with vapour from the evaporation plant down-stream effects (for example no. 4,5,6). 6.2.2.9 Hot water generation from evaporation plant vapor extraction Another option for increasing the hot water generation from the evaporation plant is to install a separate vapour extraction on the evaporation plant effects for hot water production. There are a number of evaporation plants that have been designed for production of hot water for reducing live steam consumption in the hot water system. Extracting vapour from the down-stream effects increases the steam consumption in the evaporation plant marginally, and if compared to the alterative of producing the hot water with live steam it is much more efficient. The indirect live steam consumption for producing 85°C hot water is around 0,2-0,5 MW steam heat per 1 MW produced hot water depending on the specific evaporation plant design. 6.2.2.10 Upgrading warm water in the evaporation plant surface condenser The warm water temperature exiting the evaporation plant surface condenser is dependent of the vacuum (steam temperature) after the last effect. Usually this vacuum is selected to produce 40-55ºC warm water at the mill required evaporation capacity. 45 VÄRMEFORSK A pulp and paper mill with spare capacity in the evaporation plant would increase reduce the cooling water flow and increase the temperature of the warm water produced. I higher warm water temperature would indirectly increase the hot water temperature. As an example, Stora Enso Nymölla pulp mill has spare capacity in the evaporation plant which is used for production of hot water instead of warm water. One of the three existing evaporation lines is feed with 45ºC warm water and produces 70-85ºC hot water [20]. 6.2.2.11 Adapting evaporation condensate temperature to process requirements The temperature of the stripped evaporation condensate can often be adjusted to the temperature that is needed in the position where it is used. If the stripped condensate is used in the brown stock washing, the condensate temperature is normally adjusted to 8590°C to reduce the need for live steam heating. The stripped condensate temperature can be controlled by adjusting how much of the condensate is cooled by the ingoing foul condensate. The B-condensate is normally used in the causticizing area and the temperature requirement is around 65-70°C. This condensate can also be internally preheated within the evaporation plant to reduce possible need for live steam in the causticizing area. The A-condensate does not require any stripping and should be flashed down to the temperature that is needed in the position where it is used. 6.2.2.12 Heat recovery from bleach plant effluent The alkaline and acid effluent in a pulp mill is cooled before it enters the effluent treatment plant. The heat recovery of the bleach plant effluent for semi-hot water production is today state-of-the-art for new mills with an efficient secondary heat system design. Normally a semi-hot water in the range 65-75°C can be produced, or alternatively the heat is used for the first stage heating of the returned district heating flow entering at around 45-55°C. The quantity and temperature of the acidic and alkaline effluent depends on the bleaching sequence applied in the pulp mill. The overall secondary heat balance and specific bleach plant conditions needs to be reviewed for the individual mill to be able to conclude it the alternative is interesting or not. It is important to note that bleach plant effluent coolers often are subject to scaling and this limits the degree of heat recovery that can be justified. 46 VÄRMEFORSK 6.2.2.13 Improved heat recovery from dissolving tank vent condenser 6.2.2.13.1 Improving the conventional design In newer systems the recovery boiler smelt dissolving tank vent gases are normally cooled and mixed with fresh air to be burned as combustion air. Vent gases from the mix tank pass through a separate scrubber to remove particulate carryover and reduce the moisture content by cooling. In older systems the dissolving tank vent gas is released through a separate stack. Cooling of the moist gas is not required, however, scrubbing is still required to reduce the particulate emissions. There are some different techniques for cooling the dissolving tank vent gases. The more common packed bed type condenser, and the tube type heat exchanger, both use cooling water and return warm / hot water. The dissolving tank vent gases include 400-500 MJ/tDS heat when cooled down to the required 45°C. The main part of the heat is present in the region 100-65°C (approx. 80% of the heat), and further cooling down to 45°C does not generate very much more heat. In some inefficient systems there is no heat recovery of the heat in the vent gases. This is however becoming more and more uncommon when designing new systems. The present conventional design with a single stage scrubber system is based on a circulation of scrubber water that is cooled in two heat exchangers. In the first stage, high value heat is recovered for hot water production, and in the second stage the scrubber liquid is cooled further with cooling water to reach the desired scrubber temperature. This is the typical design for Valmet and Andritz boilers today. The dissolving tank vent gas scrubber could also have a dedicated heat exchanger or scrubber section inserted into the gas flow to maximize the generation of high temperature heat. Such design requires a second additional heat exchanger for final cooling of the vent gas to the desired 45°C, see figure 6. Celbi mill in Portugal is presently the only existing reference of this two stage heat recovery concept. It is relevant to underline that the concept does not introduce any unproven components in the system. The principal difference to the conventional system is that there are two heat exchangers (alternatively scrubber sections) instead of one. The vent gas from the dissolving tank is a high value heat source that could serve as heating media for preheating of boiler make-up water and turbine condensate. A good system design should be capable of preheating make-up water to about 80˚C. However, to achieve this temperature it is very important that the vent gas system is design and operated to minimize leakage of infiltration air. Leakage of air into the system decreases the dew point and the potential to achieve a high temperature of produced hot water. High vacuum in the dissolving tank is the main parameter to adjust. 47 VÄRMEFORSK Figure 11: Principle description of two stage cooling of dissolving tank vent gases for maximum hot water generation. 6.2.2.13.2 Open absorption heat pumps for vent condenser A less conventional technology that seems interesting to apply on the heat recovery of the dissolving tank is hygroscopic condensation. Hygroscopic condensation is a type of open heat pump where the moisture condenses directly against a hygroscopic medium. Using this technology for recovering heat from exhaust air from paper machines and flue gases has been studied in a few projects. A pilot plant has been commissioned and operated at a paper machine at Braviken paper mill and the exhaust of a gas engine at Eskilstuna Energi [21][22]. The dew point of the wet gases in these two applications was in the temperature level of 50-65ºC. The very high dew point slightly below 100°C of the vent gas would make it possible to produce LLP steam that could be used internally in the recovery boiler building for preheating of boiler feed water. Caustic soda is a stable hygroscopic media that could be used for the vent gas scrubbing. The system could be kept relatively clean by purging some of the NaOH to the liquor cycle through the dissolving tank. This NaOH could be part of the mill´s total Na make-up. 6.2.2.14 Hot water generation from lime kiln flue gases Normally, the lime kiln flue gases exit the smoke stack at around 250ºC. This heat could be recovered in a heat exchanger installed in the flue gas duct after the electrostatic precipitator. 48 VÄRMEFORSK In Sweden there are references of lime kilns equipped with scrubbers where low grade heat (50-60ºC) is recovered for warm water production. However, often there is a surplus of heat at this temperature level and there is a low value of the recovered heat in a scrubber. A flue gas cooler, similar to what is installed on recovery boilers recently, could recover the flue gas heat at a higher and more useful temperature level. There are presently no known references within the pulp and paper mill. The technology and economic feasibility of installing a flue gas cooler on a lime kiln is studied in the ongoing Värmeforsk project 12-244. 6.2.2.15 Reduce consumption of hot water One of the more evident measures to improve the secondary heat balance is to reduce the consumption of hot water. In chemical pulp mills the main part of the produced hot water is used for washing in the bleach plant and oxygen delignification. Hot water consumption can be reduced by increasing the bleach plant closure, i.e. reuse of filtrates and avoiding unnecessary dilution, as discussed in chapter 3.2. Reduced need of hot water is not only beneficial from a water consumption and effluent volume point of view; it also brings significant positive effects to the secondary heat balance. The benefit is smaller than in the case where the heat in the bleach plant effluent is recovered for hot water generation. This simply because lower water consumption indirectly means that less heat can be generated in the effluent cooler due to lower flow. Using efficient equipment for washing reduces water consumption. The reference mill described in appendix C, D and E uses the best available technology from water consumption point of view. Existing mills often have higher water consumption with wash filters installed instead of wash presses. The main driving force for replacing old wash filters with for example wash presses is although not the potential energy saving related to the secondary heat system. For paper machines and pulp machines the hot water box before the press section consumes 1-2 m3/ADt. Heating the pulp web to improve the machine press section performance can be performed with a steam box instead. Each individual case needs to be analysed to determine if there is a net benefit from energy point of view to use steam as primary heat instead of produced hot water. Generally, a steam box enables higher dryness after the press, and from a quality point of view it is easier to control the dryness profile. Hot water is often used for chemical preparation. The consumption is although relatively low and is difficult to reduce because it is determined by the concentration at which the chemical is used. 49 VÄRMEFORSK 6.2.3 Utilization of high temperature secondary heat 6.2.3.1 Reduce cooling of weak black liquor The alternative of increasing the hot water generation from cooling the WBL is discussed in chapter 6.2.2.8. It was concluded that it could be beneficial to increase the cooling of the WBL if there is a hot water deficit in the mill even though the evaporation steam consumption increases somewhat. The opposite relation is valid if there is an excess of hot water/ secondary heat in the mill. The WBL cooling could in this situation be reduced and the evaporation plant fed with higher liquor temperature resulting in reduced steam consumption. There are however practical limitations that restrict the feed temperature of the WBL. Primarily the risk of flashing in the atmospheric WBL storage tank to the DNCG system limits the temperature to around 97°C. 6.2.3.2 Hot weak black liquor flash in evaporation plant This alternative is principally the same as option 6.2.3.1with reduced cooling of weak black liquor. The difference is that the weak black liquor produced in the digester is never cooled and enters the evaporation plant at around 110°C. This requires a somewhat modified evaporation plant where the weak black liquor is flashed to a separate section of the evaporation effects. The system must also be designed considering CNCG and the separation of turpentine. Sending hot weak black liquor to the evaporation plant would eliminate the significant hot water production in the digesting plant and this option should only be considered if there is a very large surplus of hot water. The concept is expected to mainly be relevant for mills producing unbleached pulp with limited hot water consumption, or dissolving pulp with low cooking yield and proportionally more black liquor per produced ton of pulp. Richards Bay, South Africa is one of the few references of this concept. The alternative has been discussed for new mills but has so for not been included. The main reason is few existing references and the fact that there is typically a hot water deficit in a new pulp mill and should therefore not be considered. It is also foreseen that the operation of the evaporation plant would be more challenging when it is more closely interconnected with the operation of the digester producing the hot liquor flash directly into the downstream effects. 6.2.3.3 Preheating of WBL in evaporation plant Preheating of black liquor within the evaporation plant has been studied but there are no known references. The principle is to use excess hot water at around 85-90°C for preheating the black liquor in-between the last and second last evaporation effect. 50 VÄRMEFORSK The primary heat saving in first effect would however be very marginal. A seven effect evaporation plant has a theoretical steam economy of 7 ton evaporated per ton consumed steam. 6.2.3.4 Preheating chemicals used in bleaching Chemicals used for pulp bleaching can be preheated. Specifically chlorine dioxide (ClO2) is added as a low concentration solution. The ClO2 solution leaves the chlorine dioxide plant absorption with a temperature of 8-15°C and enters the D-stage at 7590°C. Preheating the ClO2-solution to 50°C in a titanium heat exchanger with for example bleach plant effluent reduces the steam consumption in the bleach stage by up to 80 kg/ADt at a ClO2 flow corresponding to 1 m3/ADt. Other chemicals used for bleaching are fairly concentrated, or do not accept handling at high temperature due to safety concerns or risk of decomposing. This limits preheating of all other chemicals than chlorine dioxide. 6.2.3.5 Integration with district heating net The Swedish pulp and paper industry sells about 2.0 TWh/year district heating per year whereof 1.2 TWh/year is recovered as secondary heat. In year 2011, 22 mills sold heat to a district heating system [7]. The delivery of secondary heat to the district heating net is although limited. First of all the quantity of secondary heat is limited by the district heating return temperature as well as the outlet temperature required by the consumers. A well designed district heating network with both low outlet temperature and low return temperature makes it easier to find secondary heat sources that have high enough temperature. The district heating return temperature can be in the range 40-70°C and the outlet temperature 80-130°C. For a mill that is already interconnected the best way to deliver more heat is to increase the temperature of the available secondary heat. The district heating return is heated in a first stage with this secondary heat. This considers optimizing the hot water system for generation of hotter hot water. It may also be relevant to dedicate one of the hot water producers with extra high temperature for the purpose of delivering primarily to the district heating net. Higher value sources of heat such as flue gas coolers and vapour extractions in the evaporation plant are used for top heating of the delivered district heating water. Lower value heat from possible bleach plant effluent coolers or the paper machine/pulp dryer heat recovery is used as an initial stage for heating the district heat return. 51 VÄRMEFORSK 6.2.3.6 Integration with biomass drying Excess secondary heat can be utilized for external processes such as drying of biomass. Presently there are two mills in Sweden that have installed biomass dryers, Södra Cell Mönsterås and Södra Cell Värö. Making use of excess secondary heat reduces the operating cost compared to using primary heat for drying. The available temperature level of the excess hot water is very important since it affects the capacity of an existing dryer, and the required size of a new dryer. Even though the cost for primary heat can be reduced significantly by using secondary heat, the investment and other operating costs such as electric power, and maintenance needs to be paid with revenues/savings related to the dried biomass. The key to reducing the biomass dryer investment cost and electric power consumption is to make as high as possible secondary heat temperature available for the dryer. Both the electric power consumption and investment cost decrease with the increasing temperature of the secondary heat source. Figure 12: Relative investment cost indication for belt biomass dryers at different hot water feed temperatures. 6.2.3.7 Organic Rankine cycle (ORC) There are two market pulp mills that have experience of supplying secondary heat to an organic rankine cycle. These mills are Skutskär mill and Aspa mill. A third party (Opcon) owns the equipment that is used for producing electric power from the secondary heat. Principally the equipment consists of a special type of turbine that is 52 VÄRMEFORSK designed for operating at low temperatures and with an organic media instead of steam/water. The thermal efficiency depends on the temperature of the secondary heat (higher temperature gives higher efficiency) and the available cooling temperature (lower cooling water temperature increases efficiency). The efficiency is although generally rather low and the process consume cooling water. The feasibility of utilizing surplus heat for power generation in ORC is not studied further in this report. 6.2.3.8 Optimal preheating of demineralized feed water and steam condensate for boilers Preheating of boiler make-up water is generally applied an all Swedish pulp and paper mills. The temperature to which the boiler make-up is heated is however different from mill to mill. The make-up water that comes from the demineralization plant typically has a temperature of 25°C and for market pulp mills the target should be to preheat the makeup water to around 80°C before it enters the feed water tank. In the feed water tank steam heat is used for combined deaeration and heating before the water enters the boiler. Sending the boiler make-up to the feed water tank at for example 80°C instead of 25°C reduces the steam consumption in the deaerator and results in an energy saving. The target should be to preheat the boiler feed water as much as possible and the practical limit is set by at which temperature level where there is no surplus secondary heat is available. An adequate temperature difference needs to be maintained between the feed water tank and incoming water to allow sufficient deaeration. For integrated and non-integrated mills the achievable temperature is somewhat lower than for market pulp mills. This since the secondary heat surplus is generally lower in these types of mills. Typically a well-designed system can achieve 50-65°C with secondary heat. Figure 13: Convectional boiler treatment design and preheating with secondary heat 53 VÄRMEFORSK Boiler feed water treatment systems have evolved during recent years in conjunction with higher feed water quality requirement due to higher boiler steam temperature and pressure. A modern system includes also polishing of the process steam condensate together with the produced make-up in mixed bed filters. Mixed bed filters must operate at a relatively low temperature so the hot process condensate needs to be cooled before treatment. This requires a different design of the system to allow efficient feed water treatment combined with preheating. The main principle is to divide the treated water to the feed water tank into two streams whereof one is used for cooling the incoming process condensate and the other is heated with secondary heat. Figure 14: Modern boiler treatment design and preheating with secondary heat 54 VÄRMEFORSK The secondary heat source used for prehating boiler feed water varies. Some of the most recent recovery boilers are equipped with flue gas coolers used for preheating boiler feed water. There are several references where the dissolving tank vent heat recovery is integrated with the boiler water preheating. Many of the older mills used excess hot evaporation condensate or hot water for preheating. The most advanced system discussed for the latest green field projects is extensive preheating of the make-up water through vapour extractions in the evaporation plant. 55 VÄRMEFORSK 6.3 Alternatives for reduced cooling water consumption and related energy consumption 6.3.1 Principles to minimize cooling water consumption The main principles for minimizing cooling water consumption can be summarized as follows: 1. Minimize primary heat consumption because it is directly related to the cooling requirement of the process. 2. Avoid recovery of low grade heat from vapours and flue gases if there is a surplus of secondary heat at the same temperature level. 3. Increase the delta T (difference between ingoing and outgoing temperature) of cooling water in order to lower cooling water flow. 4. Increase cooling of the process through delivery of secondary heat to consumers outside the plant 5. Install equipment to enable evaporative cooling to further minimize water consumption for cooling purposes. These five main principles and related measures are commented under chapters 6.3.26.3.5. 6.3.2 Primary heat consumption The cooling requirement of the process is directly related to the amount of primary heat introduced to the system as the different fuels to the boilers. The key to reducing the cooling requirement is therefore to minimize primary heat consumption to a minimum. This can be accomplished by general energy conservation (i.e. higher steam economy in evaporation, modern heat recovery on digesters, efficient secondary heat recovery with minimized steam usage for water heating etc.). Reducing specific heat consumption will reduce the heat that needs to be cooled away. Pulp and paper mills equipped with cooling towers will experience better operating margins in the cooling towers with lower temperature of the distributed cooling water. In a typical Nordic mill, a lower cooling requirement translates into a reduced raw water consumption since less cooling water is needed. Accordingly, a large part of the cooling water issue is integrated with the mill primary heat consumption (GJ/ADt). This is often not recognized and is not included in the overall justifications for primary heat conservation measures. On the other hand, reducing the primary heat consumption can seldom be justified only by the reduction in cooling water requirement. As comparison, the cost of supplying cooling water is estimated within this study to about 0.5 SEK/MWh for mechanically treated water (see Appendix B). This can be compared to the LP steam cost of around 127 SEK/MWh (see Appendix B). 56 VÄRMEFORSK Options for energy conservation focused only on minimizing the process need of primary heat are not discussed in detail in this report. Optimization of secondary heat recovery and resulting reduction of primary heat usage are discussed in chapter 6.2. 6.3.3 Avoid recovery of low grade heat from vapours Besides reducing the heat input to the system, it can sometimes be relevant to reduce the heat recovery from the process. It is always important to have a high efficiency of heat recovery in boilers and other processes where primary heat is produced. However, when it comes to secondary heat present in vapours it can be different depending on the temperature level of the source and internal need for low temperature heat. Recovery of water vapours through condensation is relatively common in scrubbers on boilers and kilns (flue gas condensation) as well as on drying machines and paper machines. Generally it is difficult to recover this heat at temperatures above 65°C. The maximum temperature is set by the dew point for the drying air or flue gas. Heat pumps could theoretically upgrade the heat to a higher temperature level but so far there are very few references of heat pumps applied on wet air in operation in the pulp and paper industry. If there is a surplus of hot water above 65°C it becomes questionable if it is relevant to recover heat from moist air. Condensation of the exit drying air or flue gas only increases the cooling requirement of the process. It is better to allow the waste heat to escape from the system as latent heat in the water vapour (as is done in a cooling tower). Improving the secondary heat system to avoid unnecessary condensation of heat in wet air could in this context have benefits from two points of view, reduced cooling water consumption and reduced need of primary heat. Avoiding low grade recovery of heat from wet air is more interesting for market pulp mills that typically have higher surplus of secondary heat above 65°C. Integrated and non-integrated paper mills are more dependent on the heat recovery from wet air on the paper machines. Mills which have an excess of secondary heat below 65°C are recommended not to recover heat from drying air or scrubbers on boilers because it increases the cooling requirement in the process. 6.3.4 Minimize cooling water flow by increasing delta T For Nordic mills that use raw water as a cooling source there is a second aspect that needs to be taken into account when targeting lower cooling water consumption. Besides the “MW:s” that need to be cooled away, the delta T of the cooling water 57 VÄRMEFORSK determines the number of m3/ADt that is needed. In other words, if the cooling water enters at lower temperature and/or exits at higher temperature, the m3/s is reduced with the same cooling demand in MW. Generally, this requires careful review of heat exchanger capacities, and control strategy of coolers for each cooling position. Increasing the delta T, and reducing the cooling water flow, will as a consequence reduce the power consumption for pumping water. Normally the cooling water pumps operate at fixed head around 5 bar, and any reduced pressure drop because of lower flow in one of several cooling circuits, will not allow a reduction to the pump head. In reality the power saving in the distribution pump will be limited to a lower volumetric flow through the pumps. The possibility of increasing the delta T (difference between ingoing and outgoing cooling water temperature) of cooling water is analysed further in chapter 7.3 and appendix J. Undersized heat exchangers for cooling the process need a higher cooling water flow to improve the delta T. Generally, if the observed delta T in existing heat exchangers is below 10°C, the cooling water flow is too high and the heat exchanger is too small. In this case it is relevant to complement the existing heat exchanger with a second unit. New coolers are recommended to be sized with a delta T (deference between ingoing and outgoing cooling water temperature) based on design recommendations for new coolers discussed in chapter 7.3 and Appendix J. 6.3.5 Cooling by delivery of secondary heat to external consumers The alternative of increasing the delivery of secondary heat to external clients is discussed in chapter 6.2.3. The possibility of delivering secondary heat and indirectly reducing the cooling demand through district heating is discussed in 6.2.3.5. The possibility of utilizing secondary heat in biomass dryers is commented in chapter 6.2.3.6. In this case the heat leaves the system as latent heat escaping the biomass dryer. External organic rankine cycles (ORC) for power generation from low grade heat are discussed in chapter 6.2.3.7. Heat could also be delivered to external clients such as saw mills and other industries. However, the temperature requirement is often relatively high (>100°C) and it is difficult to find excess secondary heat at this level. 58 VÄRMEFORSK 6.3.6 Cooling water system with cooling towers The most obvious measure to reduce cooling water consumption is to install cooling towers. The amount of heat that needs to be removed from the process is unchanged but it leaves the system as vapour instead of heated water. There are several ways that cooling towers can be integrated. For new green field mills it is common to size the cooling towers with sufficient capacity to cover the whole cooling demand of the mill. The reference mills discussed in appendix C, D and E. include a cooling tower concept for all used cooling water except hydraulic coolers. In Sweden there is at present only one mill that has been designed with cooling towers. This is the Mönsterås mill. There are however several mills that have cooling towers installed for cooling the effluent from the fiber line to the secondary treatment. The effluent is passed directly thought the cooling tower. There are also mills that cool the effluent indirectly with cooling water in heat exchangers and this has now been the more preferred solution because of the increasing focus on odour reduction and minimization of risk for legionella. It is however important to note that heat exchangers fed with cooling water have the drawback of increasing cooling water consumption. Cooling towers have slightly higher power consumption compared to using fresh water for cooling. Generally, a cooling tower system should be avoided if sufficient quantity of fresh water is available during all seasons of the year. Also the recipient should be capable of receiving the thermal load of the exiting cooling water. Another drawback with cooling towers is the visual impression of the cooling tower plume. This is no environmental problem; but water vapours can be associated with air pollution from boiler stacks. 59 VÄRMEFORSK 7 Evaluation of selected alternatives Some of the measures discussed in chapter 3 to 6 have been evaluated further in this chapter. Cost related to energy, water and chemical consumption is included in Appendix K. 7.1 Selection of energy and water conservation measures A number of process water conservation options are discussed in chapter 3,4 and 5 for respective mill type. The focus is on defining measures for the modern reference mills. All of the discussed process water conservation options are summarised in Table 7.1 below. The order of magnitude potential water saving has been approximated where possible. Some of the alternatives have been selected for further review of the technical and economic feasibility. Proven technology, preliminary judged economic potential and magnitude of water saving is the bases for the selection. 60 VÄRMEFORSK Table 7.1: Summary of process water conservation options Report chapter 3.2 3.3 3.4 Process water conservation measure Further closure of the reference mill bleach plant Return acid filtrates from bleaching to liquor cycle Separate evaporation and incineration of acid filtrates Nonintegrated bleached kraft market pulp Integrated production of unbleached kraft pulp and liner Integrated production of mechanical pulp and magazine paper Potential water cons. related to ref. mill X - - <5 m3/ADt YES X - - <9 m3/ADt NO X - - <9 m3/ADt NO X X <22 m3/ADt NO X X 9 NO X X <4.6 m3/ADt10 NO X X X <1 m3/ADt NO X X X <0.5 m3/ADt NO X X X <0.2 m3/ADt NO X X X <1 m3/ADt NO - - X <0.3 m3/ADt NO - - X <4 m3/ADt NO - - X <0.3 m3/ADt NO - - X <5 m3/ADt NO - - X <6 m3/ADt NO - - X <6 m3/ADt YES Selected for further evaluation from bleaching 3.5 3.6 3.7 3.8 3.9 3.10 3.11 5 5 5 Impregnation of chips for uniform moisture content before refining with white water instead of fresh water. 5.1 5.2 5.3 9 Partial recovery of process effluent to water treatment plant or selected positions in the process Membrane filtration of effluents and internal reuse Alternative usage of evaporation condensates Reduced consumption and recovery of sealing water Recovery of purge from cooling towers Collection of steam condensate from pipe racks Returning cooling water from oil coolers Substitute fresh water used in chemical preparation with super clarified filtrate from paper machine Substitute fresh water used on the paper machine high pressure showers with super clarified filtrate. Membrane filtration of effluents and internal reuse Recovery of effluent to water treatment or selected positions in the process Evaporation of effluent Difficult to quantify order of potential saving. Figure valid for the kraftliner reference mill. The reference mill for market pulp use 100% condensate. 10 61 VÄRMEFORSK Table 7.2, Table 7.3 and Table 7.4 summarize the secondary heat and cooling water optimization measures. The potential energy saving and utilization has been quantified. The most technically proven measures with high potential energy saving has been economically evaluated in chapter 7.3. 62 VÄRMEFORSK Table 7.2: Summary of heat recovery measures Report chapter 6.2.2.1 6.2.2.2 6.2.2.3 6.2.2.4 6.2.2.5 6.2.2.6 6.2.2.7 6.2.2.8 6.2.2.9 6.2.2.11 6.2.2.12 6.2.2.13 6.2.2.14 6.2.2.15 11 Heat recovery measure Review sizing of heat exchangers Cascade interconnection of heat exchanger recovery units Monitoring heat exchanger performance Monitoring secondary heat system performance Preheating of fresh water Hot water temperature and size of hot water tank Storage of energy in hot water accumulator Increased hot water generation from weak black liquor Hot water generation from evaporation plant vapor extraction Adapting evaporation condensate temperature to process requirements Heat recovery from bleach plant effluent Improved heat recovery from dissolving tank vent condenser Hot water generation from lime kiln flue gases Reduce consumption of hot water Nonintegrated bleached kraft market pulp Integrated production of unbleached kraft pulp and liner Integrated production of mechanical pulp and magazine paper Potential heat gene. related to ref. mill (as heat) X X X 11 YES X X X 11 NO X X X 11 NO X X X 11 NO X X X <0.1 GJ/ADt NO NO Selected for further evaluation X X X 11 X X X 11 NO X X - <0.5 GJ/ADt NO X X - <4 GJ/ADt NO X X - <0.5 GJ/ADt NO X X X <1 GJ/ADt NO X X - <0.4 GJ/ADt YES X X - <0.2 GJ/ADt NO X X X <0.5 GJ/ADt NO Difficult to quantify potential 63 VÄRMEFORSK Table 7.3: Summary of heat utilization measures Report chapter 6.2.3.1 6.2.3.2 6.2.3.3 6.2.3.4 6.2.3.5 6.2.3.6 6.2.3.7 6.2.3.8 Heat utilization measure Reduce cooling of weak black liquor Hot weak black liquor flash in evaporation plant Preheating of WBL in evaporation plant Preheating chemicals used in bleaching Integration with district heating Integration with biomass drying Organic Rankine cycle (ORC) Optimal preheating of demineralized feed water and steam condensate for boilers Nonintegrated bleached kraft market pulp Integrated production of unbleached kraft pup and liner Integrated production of mechanical pulp and magazine paper Potential heat utalization related to ref. mill (as prim heat) Selected for further evaluation X X - <0.5 GJ/ADt NO X X - <0.5 GJ/ADt NO X X - <0.3 GJ/ADt NO X X X <0.2 GJ/ADt NO X X X 12 NO X X X <0.7 GJ/ADt13 NO X X X ∞ NO X X X <0.4 GJ/ADt NO Potential heat gene. related to ref. mill < 30 kWh/ton NO Table 7.4: Summary of cooling water conservation options Report chapter 6.3.2 6.3.3 6.3.4 6.3.5 6.3.6 Cooling water conservation measure Primary heat consumption Avoid recovery of low grade heat from vapours Minimize cooling water flow by increasing delta T Cooling by delivery of secondary heat to external consumers Cooling water system with cooling towers Nonintegrated bleached kraft market pulp Integrated production of unbleached kraft pup and liner X X Integrated production of mechanical pulp and magazine paper X X X X < 5 kWh NO X X X < 20 kWh/ton YES X X X 12 NO X X X - 20 kWh/ton YES Selected for further evaluation (as power) Above potential related to cooling water consumption is very indicative. As comparison the total power consumption related to raw water distribution, cooling towers and effluent treatment is 70 kWh/ADt for the market pulp reference mill. 12 13 Difficult to quantify potential. Dependent on the specific location of mill. Based on drying of all internally generated bark to 60% dryness. 64 VÄRMEFORSK 7.2 Evaluation of selected process water conservation options 7.2.1 Increased closure of the reference mill bleach plant Alternatives of reducing the bleach plant water consumption and process effluent in the reference market pulp mill is discussed in chapter 3.2. The difference in operating economy of the four defined alternatives compared with the reference mill has been estimated: Table 7.5: Estimated effects on operation cost with increased closure of bleach plant Case A Case A with Cl removal Case B Case B with Cl removal Chemical costs NaOH to white liquor CaO to lime kiln SEK/ADt SEK/ADt -3.4 -1.9 -1.5 -1.9 -9.6 -0.4 -6.6 -0.4 Increased ClO2 usage SEK/ADt 1.9 1.9 1.8 1.8 Increased O2 usage SEK/ADt 0.1 0.1 0.2 0.2 Increased H2O2 usage SEK/ADt 0.8 0.8 1.6 1.6 H2SO4 (usage in Cl/K kidney) (100%) SEK/ADt 0.0 0.0 0.0 0.0 Process water Fresh water cost Effluent treatment cost SEK/ADt SEK/ADt -3.5 -4.0 -3.5 -4.0 -3.5 -3.9 -3.5 -3.9 Energy Fuel consumption in lime kiln Heat generation in recovery boiler Evaporation plant steam consumption Fiber line steam consumption Total mill power consumption SEK/ADt SEK/ADt SEK/ADt SEK/ADt SEK/ADt 0.6 0.6 0.0 0.0 0.0 0.5 0.6 0.0 0.0 0.0 0.7 0.9 0.0 0.0 0.0 0.5 0.9 0.0 0.0 0.0 Summary Net saving Annual saving (2000 ADt/d) Total estimated investment cost whereof CRP as installed Simplified pay-back SEK/ADt MSEK/y MSEK MSEK Years 8.7 5.7 5 0 0.9 7.0 4.5 38 33 8.3 12.1 7.9 5 0 0.6 9.3 6.1 50 45 8.2 The fresh water and effluent reduction corresponds to 5 m3/ADt. The COD reduction of 1.0-1.6 kg/ADt going to primary treatments is not as significant. Process key data is presented in chapter 3.2.3. 65 VÄRMEFORSK The operating cost does not include additional chemicals for neutralizing the bleach plant effluent (NaOH or CaO). The total investment cost includes an allowance for an additional buffer tank for P-filtrate and related piping. Simulations indicate no increased risk for scaling related to NPE and no reduced availability is therefore accounted. The improved operation cost justifies approximately 40-60 hours extended downtime per year (0.5% availability). One day lost pulp production in the reference mill equals approximately 4 MSEK. The investment cost includes total cost for Cl removal process where applicable. An allowance of 5 MSEK has been included for piping modifications for enabling operation with white water and condensate usage as recommended in addition to the reference mill concept. Increased integration of process flow within the bleach plant may imply that larger buffer sizes are required to maintain the runnability. Additional cost and allowance for this has not been included, however for the reference mill size it is in the several million SEK. 7.2.2 Evaporation of process effluent in the integrated magazine paper mill Increased closure of the magazine paper mill by evaporation of process effluent and recirculation of condensate to the TMP pulp mill has been evaluated further. The surplus steam available in the reference mill allows evaporation of 65% of the produced process effluent and the fresh water consumption and can be reduced further. The remaining fresh water is used on the paper machine to maintain high product quality. Table 7.6: Key data for evaluation Paper mill data Paper production (average) Reference mill effluent flow effluent flow (average) effluent COD (to primary) Surplus steam Ditto [ton/d] [m3/ton] [m3/h] ton/d [GJ/ton] [MW] Evaporation plant (6 effects) Part of effluent sent to evaporation Steam temperature Condenser temperature Evaporation heat economy Power consumption Inlet dryness [%] [°C] [°C] [MJ/m3 evap] [kWh/m3 evap] [%] 66 1284 5.8 310 52 1.45 21.5 65 135 80 400 3 0.7 VÄRMEFORSK Outlet dryness Evaporated Hot water generation (50-80ºC) Clean condensate ditto average Cooling water flow 15-45ºC (if not integrated to dryer) [%] [t/h] [m3/h] [m3/ton] [m3/h] 15 124 140 3.7 200 [m3/h] 160 Biomass dryer Biomass flow Inlet dryness Outlet dryness Evaporated Hot water consumption (80-50ºC) ditto [tDS/h] [%] [%] [ton/h] [m3/h] [MW] 16.2 45 60 9 120 10 Operation cost evaporation plant14 ditto specific operating cost Investment cost evaporation plant ditto specific investment cost (description 10 years) [MSEK/year] [SEK/m3] [MSEK] [SEK/m3] 0+2+1=3 1.8 ≈ 200 11.7 COD reduction to recipient with 80% reduction in treatment plant (as flow to primary) Process water saving in paper mill [kg COD/ton] [m3/ton] 5.2 (26) 3.7 Calculation assumes that process effluent is taken before entering the primary treatment. The concept could possibly be improved further by selectively separating high COD streams for evaporation. Pretreatment to increase concentration of effluent with membranes as discussed in chapter 3.6 may also reduce the total cost. If the evaporation plant is integrated with a biomass dryer, less cooling water is required for the process at steady-state operation. Fresh water is used as cooling media as backup or continuously if no biomass dryer is integrated. No cost for disposal of the produced liquor is included in the calculation. If disposed in a nearby chemical pulp mill it can be assumed that the energy cost is low compared to the capital cost. The increased steam consumption in the chemical pulp mill evaporation plant is compensated by increased steam generation in the recovery boiler. Disposal in a chemical pulp mill requires spare capacity in the recovery boiler and evaporation plant. 14 Surplus steam cost is set to 0 SEK/MWh. Electric power cost estimated as 3 kWh/ton evap. Allowance for maintenance and manning is set to 2% of investment cost and 1 MSEK/year, respectively. 67 VÄRMEFORSK Incineration in a power boiler or dedicated furnace on site requires higher dryness of the produced liquor. Significant increase in investment cost related to the evaporation as well as equipment for enabling the incineration is then required. 7.3 Evaluation of selected secondary heat improvements and cooling water conservation options 7.3.1 Two stage heat recovery of dissolving tank vent gas The alternative design with a two stage heat recovery from the dissolving tank vent gases is discussed in chapter 6.2.2.13. The economic feasibility of a two stage design instead of a single stage for the reference market pulp or kraftliner mill is estimated below. The market pulp reference mill described in appendix C has been selected as basis for the recovery boiler size. Table 7.7: Key data for evaluation Parameter Recovery boiler load Heat in vent gas (100°C to 42°C) ditto Percent recovered in first stage ditto Percent recovered in second stage ditto Temperature of water to first stage Temperature of water to second stage [tDS/d] [MJ/tDS] [MW] [%] [MW] [%] [MW] [°C] [°C] Temperature from first stage (heat recovery stage) [°C] Temperature from second stage (cooling stage) Produced hot water flow Cooling water flow (second stage) Heat flow (ref 0°C) where-of (65-80C) Heat exchanger relative size Investment cost [°C] [kg/s] [kg/s] [MW] [MW] [-] [MSEK] Value of heat (65-80°C) as LP steam 150 SEK/MWh Simplified Pay-back [MSEK/year] [years] Conventional Two stage one stage heat recovery 3477 3477 440 440 17.7 17.7 100 85 17.7 15.1 0 15 0.0 2.7 40 65 15 80 105 35.4 6.6 1.0 ref ref 80 30 239 42 80.3 15.1 1.7 3.6 10.7 0.3 The performance and design data must be calculated for each individual case. The above information is indicative for a mill with a deficit of secondary heat in the range 65-80°C. 68 VÄRMEFORSK The one stage system cannot be feed with water warmer than about 40º and still maintain sufficiently low moisture content of the vent gas. This limits the amount of 80ºC that can be produced in a single stage system where dissolving vent gas is used as combustion air in a recovery boiler. The two stage heat recovery is feed with cooling water for guaranteeing the wet gas exit temperature (≈ 45ºC dew point), and semi-hot water of 65ºC for maximizing hot water production. 7.3.2 Installing cooling towers The most apparent measure to reduce the cooling water consumption is to install cooling towers. The capital cost and operating cost for this alternative has been estimated for a 50 MW cooling tower substituting usage of mechanically treated raw water for cooling purposes. Table 7.8: Key process and cost data related to a typical cooling tower installation. Parameter Cooling tower load Cooling water flow Inlet temperature Outlet temperature Power consumption cooling tower fans Power consumption pumps in distribution pumps Unit MW m3/h °C °C kW kW Data 50 1714 45 20 132 47 Investment cost cooling tower Total investment cost incl. erection Capital cost (description set to10 years) MSEK MSEK MSEK/year 6 21 2,1 Operating cost power Savings related to reduced water consumption Miscellaneous other costs Total operating cost MSEK/year MSEK/year MSEK/year MSEK/year 0,7 -0,3 0,4 0,8 Total cost including capital MSEK/year 2,9 The investment cost and operation cost varies between plants depending on the specific conditions. Above data should be regarded as an order of magnitude cost indication. There is no economical return on the investment in cooling towers since the operating cost exceeds the saving in the raw water treatment plant. Possibly avoided investment in the raw water treatment to secure the raw water intake capacity is not included in the evaluation. This could be accounted if the mill raw water consumption is increased and the cooling tower investment makes it possible to avoid an upgrade of the water treatment plant. This saving would although be almost 69 VÄRMEFORSK negligible compared to the cooling tower investment cost (say in the order of 20% of the cooling tower investment). The majority of Swedish pulp and paper mills are not limited on the raw water treatment capacity or by limited fresh water availability (see Appendix C) The specific cost for saving water by installing cooling towers is in the order for 0.2 SEK per saved m3 raw water. There is a clear environmental benefit of installing cooling tower if the fresh water is limited as it can be in South Europe during the dry seasons of the year. The thermal load on the water recipient can be avoided with cooling towers and this can also be of importance for mills with relatively small recipient. In Scandinavia this is seldom the case and the higher energy consumption of the cooling towers, compared to using fresh water for cooling, would have a net negative impact on the environment. 7.3.3 Optimal sizing of new heat recovery units 7.3.3.1 Design recommendations The question of how far it is economically feasible to drive the heat recovery is often raised when designing new heat recovery systems or when a new unit is added to an existing unit. The following report chapter and Appendix I presents a general method of determining the optimal temperature difference for sizing of heat exchangers. The main conclusion from the analysis is that heat recovery units should be generously sized considering today’s high energy cost of primary heat in Swedish pulp and paper mills. Typical cost level for heat exchangers results in a theoretic optimal sizing corresponding to a minimum temperature difference of 1-9°C between the two medias that is heat exchanged. It is important to consider non-ideal counter current conditions (channelling in heat exchangers) when a very low temperature difference can be justified. The optimal temperature difference ΔT depends on the ratio between the heat capacity (kJ/°C) of the cold and hot flow that is heat exchanged. The optimal temperature difference for a given heat exchanger cost is highest when the cold and hot flow have the same heat capacity, i.e. ratio is 1. 7.3.3.2 Design chart The theoretical optimal design point at maximum continuous rate (MCR) has been plotted for a “base case” representing typical marginal heat exchanger cost and value of primary energy. The assumed process and cost data is presented in appendix I. 70 VÄRMEFORSK The optimal design point has also been calculated for two additional cases representing double value of primary heat or half equipment cost, as well as half value of primary heat or double equipment cost. Selecting a design point on any of these two design lines is analogous to doubling or halving any of the above base case assumed design parameters. The optimal temperature difference (y-axis) is plotted against the logarithmic scale of the heat capacity ratio (x-axis) in the diagram presented below. Figure 15: Design diagram for heat recovery units As an example, a heat exchanger designed for conditions equivalent to the base case and with a heat capacity (kJ/°C) ratio of 2 yields an optimal temperature difference of 2.7°C. The analytical solution to the above design chart is also presented in appendix I. 71 VÄRMEFORSK 7.3.4 Optimal sizing of new coolers 7.3.4.1 Design recommendations When replacing or installing new heat exchangers for only cooling purposes it is relevant to review what is a suitable sizing is taking into account both the investment cost, but also the water consumption and related operational cost in the raw water treatment and distribution. The optimal design point of coolers differs significantly compared to heat recovery units A large heat exchanger would have a lower cooling water consumption and operation cost, but the capital cost is high. A small heat exchanger would have high cooling water consumption and power cost for pumping water, but low capital cost. Therefore, the suitable sizing will depend on the power price and power consumption, relative the investment cost for the heat exchanger installation and the pay-back criteria that is applied. To determine the suitable sizing of new coolers, the total cost including both capital and operation cost has been calculated for different heat exchanger sizes in appendix J. The recommended design for a new cooler does not necessarily correspond to the calculated economic optimal design. It can be observed that the total cost increases only marginally (≈10%) when doubling the temperature increase of the cooling water. However, the size and capacity of the cooler increase by approximately 25% which gives useful design margins for future higher production. Also the design margins for upset conditions needs to be reviewed when designing a new cooler. Selecting a heat exchanger sized based on the cooling water temperature increase at the economic optimal design at MCR multiplied by 2 as recommended above is a good rule of thumb that gives appropriate design margins for most installations. 7.3.4.2 Design chart The optimal sizing of new coolers is dependent of the available cooling water temperature and the temperature of the heat source that is cooled. The initial temperature difference between the cooling water and heat source is proportional to the suitable temperature increase of the cooling water (Tout-Tin). The optimal temperature increase has been calculated for a number of temperatures differences between cooling water and the heat source (Tout-Tin). Process conditions and costs have been established for a “base case”. The assume process and cost data is presented in appendix J. The optimal design point has also been calculated for two additional cases representing double value of primary heat or half equipment cost, as well as half value of primary 72 VÄRMEFORSK heat or double equipment cost. Selecting a design point on any of these two design lines is analogous to doubling or halving any of the above presented design parameters. Figure 16: Design chart for coolers The above optimal design points has been determined assuming the heat source is cooled 10°C. Any deviation from this assumption affects the optimal design point only marginally and no correction factor has been developed. As an example, effluent should be cooled from 65°C to 45°C with cooling water of a temperature of 15°C. The conditions are the same as for the “base case” defined in appendix J. The initial temperature difference of the heat source and cooling water (xaxis) is 65°C-15°C = 50°C. Reading the above diagram for the base case according to the black arrow results in an optimal temperature increase of the cooling water corresponding to 14°C. The rule of thumb sizing is that the optimal temperature of 14°C is doubled to give good capacity margins for upset conditions and future high production. The suitable sizing is 28°C temperature increase of cooling water. 73 VÄRMEFORSK 7.3.5 Increase size and number of existing coolers The economic feasibility of doubling the size of coolers by installing a second heat exchanger in series or parallel is estimated. The calculations has been done specifically for coolers operating with a low rise in the cooling water temperature which can be found in existing mills where the production has been increased stepwise without improving the capacity in process coolers in the same rate. The feasibility analysis of adding an additional cooler has been done for both parallel and cascade configuration. If a higher pressure drop is accepted, a cascade configuration is beneficial from heat transfer rate point of view. Mechanically treated raw water is assumed as cooling media. The detailed calculations and results are presented in appendix J. Installing a completely new cooler can rarely be justified by the power saving. An additional or a completely new cooler can only be justified because of required cooler capacity to satisfy the process requirements, or alternatively if the mill is fresh water limited. However, when a new cooler is required, it is recommended that the sizing considers the energy and water consumption as outlined in chapter 7.3.4. The specific cost per saved m3 cooling water is estimated to be less than 0.05 SEK. This can be compared to the specific cost of 0.2 SEK/m3 calculated for cooling towers in chapter 7.3.2. 74 VÄRMEFORSK 8 Results and conclusions The results chapter summarize the energy and water consumption data for the selected mill types in subchapter 8.1. Results for selected and evaluated measures that can be applied primarily for reducing the fresh water consumption are presented in section 8.2.1. Similarly results related to measures that primary reduce the process effluent generation and improves the secondary heat system are presented in subchapter 8.2.2 and 8.2.3, respectively. 8.1 Compiled water and energy consumption data The study presents actual process and cooling water consumption for some of the main products in the Swedish pulp and paper industry. The performance of existing mills is compared with EU-BAT BREF 2013 as well as a hypothetical reference mill developed by ÅF and Innventia within several earlier research programs, updated in its latest revision 2010 and also to some extent within this work. Similarly energy performance data has been presented for extreme references available in the world. The BREF-BAT emissions and consumption numbers represents European agreement of the best-available-technology within the pulp and paper industry. The above guidelines are proposed in the latest draft dated July 2013, which also is foreseen to be accepted during 2014. The difference between an average Swedish mill and the reference mills is large when it comes to water use. The main part of the difference is related to cooling water system design. Based on reported water consumption, we still consider that Swedish mills have process water consumptions and process effluent flows in line with international standards and BAT. The reference mills are theoretical greenfield mills using best available proven technology, but not necessarily commercially possible to justify for all mill conditions. Still it should be noted that the water consumption in many existing mills still is significantly higher than in the developed reference mills and one reason for this is the reference mill being a greenfield mill. 75 VÄRMEFORSK 8.1.1 Market pulp The Swedish market pulp mills have relatively low process water and energy consumption compared to BREF BAT [1] [2]. However, compared to the reference mill, the consumption numbers are higher for the Swedish average mill. It is although difficult to compare the existing mills with the reference mill which is completely new and is based on the latest available technology. Table 8.1: Comparison of hypothetical reference mill, BREF BAT, Swedish average and lowest in operation for the market pulp mill. Reference Market pulp mill BREF BAT (draft 2013) Swedish Average 2012 Process water effluent ditto COD Cooling water effluent ditto heat load m3/ton kg/ton m3/ton GJ/ton 22 <9 25-50 7-20 15 16 7.6 16 Steam heat consumption Power consumption GJ/ton kWh/ton 8.9 600 13.7-18.4 13.9 700-800 796 8.1.2 35 15.6 42 ≈ 5.3 Lowest in operation at extreme conditions 23 2 15 16 9-10 520 Kraftliner The Swedish market kraftliner mills have relatively low process water and energy consumption compared to BREF BAT. The lowest in operation identified within this study represents extreme conditions and the degree of closure is generally far more than what can be justified for Scandinavian conditions. 15 16 Reference mill design includes cooling towers resulting in no cooling water effluent. Information not available 76 VÄRMEFORSK Table 8.2. Comparison of hypothetical reference mill, BREF BAT, Swedish average and lowest in operation for the 100% unbleached kraftliner mill. Reference BREF Kraftliner BAT (draft 2013) mill Process water effluent ditto COD Cooling water effluent ditto heat load Steam heat consumption Power consumption 8.1.3 m3/ton kg/ton m3/ton GJ/ton GJ/ton kWh/ton 13.7 4 15-4017 15 16 2.1 13.7 4 16 15 16 18 15-40? Swedish Average 2012 21 11.7 27 3.4 21 11.7 27 Lowest in operation at extreme conditions 0.4 16 15 16 0.4 16 16 Magazine Paper The reference mill is has low water consumption compared to the Swedish average as well as BREF BAT. The BREF BAT performance for magazinepaper and newspaper is not clearly defined because consumption data is separately presented for recycled paper plant, TMP production and the paper machine. Table 8.3: Comparison of hypothetical reference mill, BREF BAT and Swedish average. Process water effluent ditto COD Cooling water effluent ditto heat load m3/ton kg/ton m3/ton GJ/ton Steam heat consumption Power consumption GJ/ton kWh/ton Reference Magazine paper 5.8 0.8 BREF BAT 15 16 1.2 16 16 3.3 2410 16 16 16 16 (draft 2013) 9 to 16+ 23.519 0.9 to 4.520 Swedish Average 2012 14 4.7 53 The study cannot conclude the performance corresponding to the magazine paper mill which has the lowest consumption in the world. 17 Unbleached kraft 15 – 40 m3/ADt + RCF paper mills without deinking 1.5 – 10 m3/t+ Unbleached kraft pulp 2.5-8 m3/ton in BREF-BAT document 19 Sum of data for paper machine, TMP production and recycled paper plant. 20 BAT data is 0.9 – 4.5 kg COD/ton integrated production of paper and board from mechanical pulps 18 77 VÄRMEFORSK 8.2 Improvement measures 8.2.1 Measures for reducing the fresh water consumption A fresh water limited mill should primarily reduce the cooling water consumption. The measures for reducing the cooling water consumption can be implemented with a lower cost compared to technology for reducing process water consumption. The results from this study show the following cost relation per saved m3 fresh water: Table 8.4: Summary of results for fresh water conservation measures Report chapter 7.3.4 7.3.2 7.2.1 7.2.2 Evaluated measure Minimize cooling water flow by increasing delta T Increase size and number of existing coolers Cooling water system with cooling towers Further closure of the reference mill bleach plant Evaporation of effluent in newspaper mills Operation cost [SEK/m3] Capital cost21 [kSEK/m3] Potential saving [m3/ton] Comment < + 0.2 <2 20% reduction of cooling water? Only if fresh water limited Step 1 ≈ - 0.2 - 12 all cooling water Step 2 -1.2 10-100 5 Step 3 + 1.8 1 000 3.7 Step 3 The reference mill fresh water usage related to cooling purposes is already very low due to the introduction of cooling towers. Only a minor amount of the cooling water is leaving the mill through the hydraulic coolers. Reduced cooling water consumption by installing cooling towers would result in a specific cost of 0.2 SEK per saved m3. Optimizing the sizing of heat exchangers for cooling would have lower specific cost than installing cooling towers. A fresh water limited mill should not primary investigate process closure. The electric power consumption related to pumping cooling water can be reduced by optimizing the sizing of new coolers and improves the performance of existing units. The total power consumption related to pumping the reported 300 million m3 cooling water used within the Swedish pulp and paper industry is estimated to be less than 100 GWh [2]. A reduction by 10% based on optimizing coolers would give a marginal power saving but significant lower water consumption. Reducing the fresh water flow through optimization of existing coolers can only be recommended if the mill is limited in fresh water. However, sizing of new coolers should be optimized considering the energy consumption related to pumping large volumes of water. For further information see appendix J. 21 Specific investment cost (kSEK) calculated per volume (m3) reduced water flow on hourly basis. As example investment is 100 kSEK and water saving is 100 m3/h yields specific cost of 1 kSEK/m3 78 VÄRMEFORSK The potential in volumetric terms for reduction in fresh water consumption in existing mills is much higher in the cooling water systems compared to saving water in the production process (see Table 2.2, chapter 2.2). Also the cost per saved m3 fresh water is much lower compared to corresponding specific cost for measures reducing the process water consumption (see appendix K). The study shows that the incentive for reducing the fresh water consumption related to cooling purposes is relatively low in Sweden. Generally, few mills are limited on the fresh water availability and the production and usage cost of mechanically treated cooling water is very low. If the quantity of fresh water is limited, the sizing of process coolers should be reviewed as a first low cost step in order to reduce the cooling water quantity. As a second step for reducing consumption, or if the recipient is sensitive for thermal load related to the discharge of cooling water, cooling towers can be installed. Evaporative cooling is expensive and the energy consumption is higher compared to using fresh water for cooling. New coolers in the process should be designed generously to avoid unnecessarily high cooling water flow and power consumption related to pumping. Guidelines for sizing of coolers are presented in chapter 7.3.4 and appendix J. Investments for replacing existing coolers with the only purpose of reducing energy consumption cannot be economically justified. 79 VÄRMEFORSK 8.2.2 Measures for reducing environmental impact the process effluent flow and Increasing the closure of pulp and paper mills reduces the effluent emissions and the water consumption, and thereby also the environmental impact. In Sweden the effluent emissions is generally a higher concern compared to the fresh water consumption. The first step towards increased closure is to implement the best available technology described in the reference mills (see appendix C, D and E). However, it is not possible to directly implement this energy and water efficient technology in existing pulp and paper mills. In reality this modern technology is introduced step-by-step in conjunction with replacement of complete sub-process section. These investments, often in the range 100-1000 MSEK, are seldom justified economically based only on the improved energy and water performance. The investments are performed as a result of reinvestments due to escalating maintenance cost, environmental and safety concerns, and reduced availability. Alternatively, the investment can be justified by higher production capacity which is desirable if the market for the product is favorable. Accordingly, for existing mills the step towards the reference mill performance requires time. For completely new mills or after implementing the latest technology in a process section of older mills, further innovative process closure concepts are required. In this case, increased closure of the market pulp bleach plant may be justified based on reduced operating costs and lower environmental performance. The study identified and evaluated two concepts for reducing the effluent volume and COD impact of the reference mills. Significantly decreased use of water in the bleach plant is possible if increased closure is applied and the fresh water used in the bleach plant is replaced by condensate. The effluent volume could be decreased to 9 m3/ADt with 16.8 kg/ADt COD to effluent without any significant increase in the risk of formation of sparingly soluble salts in the bleach plant. To keep the chloride levels down in the white and black liquor precipitator dust leaching should be applied. Table 8.5: Summary of results for process and effluent conservation measures Report chapter 7.2.1 7.2.2 22 Evaluated measure Further closure of the reference mill bleach plant Evaporation of effluent in newspaper mills Operation cost [SEK/m3] Capital cost 22 [kSEK/ m3] COD reduction to prim. treat. (to recipient) [kg/ton] Potential effluent saving [m3/ton] Comment -1.2 ≈ 10-100 1-1.6 (0.2-0.3) 5 High volume red. + 1.8 1 000 26 (5.2) 3.7 High COD red. Specific investment cost (kSEK) calculated per volume (m3) reduced water flow on hourly basis. 80 VÄRMEFORSK Historically, there is experience of reduced availability when operating with a highly closed bleach plant. New technology such as ash leaching and improved evaporation plant condensate quality may enable new preconditions for closure of the bleach plant for both TCF and ECF pulp qualities. However, increasing the closure can also have negative impacts such as increased use of heat and power, chemical consumption etc. When evaluating the overall environmental impact also this needs to be considered for the specific case. The increased energy efficiency in new magazine paper mills exemplified in the reference mill results in a steam surplus that could be utilized for increased process closure through evaporation of effluent. The capital cost is still relatively high for evaporating the effluent and to produce power of the steam surplus is often more attractive economically. The reference magazine paper mill effluent can be reduced by up to 65% through evaporation with surplus steam. Total capital cost is estimated to be around 200 MSEK. The specific total cost considering both capital and operation is calculated to be 12 SEK per reduced m3 of the effluent discharge calculated for a 10 year period. It can also be concluded that achieving the extremely low process water consumption requires introduction of mill-wide systems for collection and reuse of sealing water and steam condensate drainage. The high quality fresh water availability in Sweden does however not make this a prioritized area. However, to reach the goal by Värmeforsk, to decrease the water consumption by 2%, should be possible for all mills and the related costs are manageable. 8.2.3 Measures for optimizing heat recovery and utilization in secondary heat systems The secondary heat systems differ considerable from mill to mill, also within the same product category. Secondary heat systems are complex and often it is a time consuming activity to audit and describe them. Improved recovery of secondary heat goes hand-inhand with the possibly of finding the utilization of the heat. An excess of secondary heat is as common as a deficit of secondary heat. The potential use of the secondary heat for other purposes than to save primary heat within the pulping process is significantly higher. Increasing the temperature of the surplus secondary heat is important to make alternative utilization of the heat economically competitive. The report gives examples of low capital cost measures that results in relatively high value energy savings. Optimization of secondary heat systems has a high potential for 81 VÄRMEFORSK short pay-back measures but relative to the total mill primary heat consumption the potential energy saving if improved process integration is expected to not be higher than a few percent. The energy savings for optimized sizing of heat recovery units are difficult to quantify. The saving for the two stage heat recovery on a dissolving tank vent condenser is calculated to give a good pay-back in a mill with a deficit of 80ºC hot water. Table 8.6: Summary of results for heat conservation measures Report chapter 7.3.1 7.3.3 Evaluated measure Two stage heat recovery of dissolving tank vent gas Optimal sizing of new heat recovery units Operation saving [MSEK/y] Capital cost Pay-back [kSEK/kW] [years] < +13.8 0.370 <1 - - 0-10 Comment Difficult to quantify The report shows that there is a potential to increase the generation of hot water (80ºC) from the dissolving tank vent gas. This is accomplished by splitting the condenser in two stages, one cooling stage and a second heat recovery stage feed with a semi hot water (65ºC). This solution is specifically interesting to apply on modern recovery boilers where the dissolving vent gas usually is sent to the recovery boiler. The report indicates that the reference mill recovery boiler can produce around 500 m3/h more hot water with potential to replace 15 MW primary heat for water heating or usage for delivery to a district heating network with a pay-back less than 1 year. If possible to apply on all the recovery boilers in Sweden this would correspond to 600 GWh per year. The general guidelines for sizing heat exchangers and analyzing their performance have potential to improve the degree of heat recovery in the Swedish existing mills. The total energy consumption related to heating process water in the temperature range 40-90ºC is estimated to be around 5% of the pulp and paper energy consumption corresponding to approximately 3 TWh. A proper sizing of heat exchangers and improved monitoring and follow-up is estimated to reduce this energy consumption up to 10%, i.e. 300 GWh. Several other process concepts for improving the secondary heat system have been presented in the report. Implementation of these rather mill specific measures has potential of saving up to the estimated total of up to 3 TWh estimated to be related to heating process water above. The potential total energy saving related to the measures discussed in the report is in the same order of magnitude as the objective. However, it has been difficult to find process water and secondary heat measures that can reduce electric power consumption by 325 GWh. This challenging target can be compared to the Swedish PFE-program (programmet för energieffektivisering) which has reported a reduction of 1 450 GWh [3]. 82 VÄRMEFORSK 9 Recommendations and use The report can serve as basis for further study and possible realization of bleach plant closure concepts for the market pulp mills. The presented water consumption statistics for Swedish pulp and paper mills is a possible basis for benchmarking cooling water and process effluent flows. It is although important to understand that differences in specific water consumption within each product category can vary due to differences in the production process and product quality, not only because of differences in technical standard. The aim of the study is to also serve as a “handbook” for improving and understanding the secondary heat and cooling systems in pulp and paper mills. A number of both general and specific measures for improving the systems are discussed. The report also presents a method for techno-economic design of new heat recovery units and coolers. 83 VÄRMEFORSK 10 Future studies Further process closure would require development of cost efficient kidneys for removal of NPE:s from the process. Precipitator dust leaching for removal of chloride and potassium is one example of a kidney already in industrial use. With increased closure of the mill other kidneys for NPE removal will be needed, i.e. for calcium and barium removal in the bleach plant. With the extensive use of black liquor condensate in the bleach plant as suggested in 3.1 a reliable quality control system will be needed. A failure in the condensate quality has to be discovered early before the condensate is used in the bleach plant. This becomes even more important as the degree of closure is increased and the use of condensate is moved later in the bleach plant. The results from trials with membrane filtration to produce permeates that could be used as process water need to be further evaluated, activities carried out within another Värmeforsk project. The separation of NPEs is depending on the composition of the organic components in the treated streams and the use of permeate in different process parts could be evaluated in simulations and in trials. The possibilities for improving the pulp and paper mills primary heat balances have been studied in national research programs, by universities and in the industry and great improvements can be observed from the 1970-ies until today. However, the secondary heat and cooling systems have not to the same degree been studied and optimized. The route to improve the recovery of heat from the process is by increased process integration combined with an improved understanding of the present techno-economical limitations of how far the heat recovery can be driven under different process conditions. Recommendations for further studies:  Development of new kidneys for removal of NPEs. For increased bleach plant closure mainly calcium and barium need to be removed.  Reliable quality measurement system of black liquor condensate for increased use.  The distribution of NPEs in membrane separation should be further studied.  Improved two stage heat recovery dissolving tank vent gas condensation  Flue gas heat recovery of lime kiln flue gases (ongoing in Värmeforsk project S12-244)  Further advanced preheating of boiler feed water with evaporation plant vapor extractions.  Pressurized hot water accumulator integrated with evaporation plant serving as new energy distribution network in-between present steam distribution (primary heat at 140ºC) and hot water systems (secondary heat at 80-90ºC). 84 VÄRMEFORSK 11 List of abbreviations Table 11.1: List of abbreviations Abbreviation ADt AOX BAT BOD BREF COD CNCG DNCG E05 ECF EIPPCB ESP HP LP MP MVR NF NPE ORC TOC WBL CTMP RCF TMP MCR SC D0 D1 EOP P Q Explanation Air dry ton of pulp Absorbable organic halogens Best Available Technology Biochemical oxygen demand Best Available Technology Reference Documents Chemical oxygen demand Concentrated non-condensable gas Diluted non-condensable gas Eldningsolja 5 (Swedish fuel oil quality) Elementary chlorine free European Integrated Pollution Prevention and Control Bureau Electrostatic precipitator High pressure steam Low pressure steam Medium pressure steam Mechanical vapour recompression Nano filtration Non-process element Organic rankine cycles Total organic carbon Weak black liquor Chemithermomechanical pulp Recycled Cellulose Fiber Thermomechanical pulp Maximum continuous rate Super calendered First chloride dioxide stage Second chloride dioxide stage Alkali, oxygen and peroxide stage Peroxide stage Stage for adding EDTA 85 VÄRMEFORSK 12 References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] European IPPC Bureau, Best Available Techniques (BAT) Reference Document for the Production of Pulp, Paper and Board, Final Draft July 2013 http://www.skogsindustrierna.org/branschen/branschfakta/miljo/miljodatabas (2014-05-20) http://www.energimyndigheten.se/Foretag/Energieffektivisering-iforetag/PFE/Resultat-och-utvardering/Resultat-fran-programmet/Investeringartotalt-och-per-bransch/ (2014-05-20) Stephen C. Stratton, Peter Gleadow, Pulp Mill Process Closure: A Review of Global Technology Developments and Mill Experiences in the 1990s, National Council for Air and Stream Improvement, technical bulletin no. 860, NCASI 2003 Åforsk Reference 09-193, Energy consumption in the pulp and paper industry – Model mills 2010, Bleached kraft market pulp mill, kraftliner, magazine paper and integrated fine paper. Berglin N, Lovell A, Delin L, Törmälä J, “The 2010 Reference Mill for Kraft Market Pulp”, TAPPI PEERS Conference, Portland, Oregon, October 2-5, 2011. Naturvårdsverket (Swedish EPA) reports 3925 and 5154 Rolf Wiberg, Magnus Forslund, Energiförbrukning i massa- och pappersindustrin 2011, Skogsindustrierna, 2012. Jonsson T, Radestrom R, Tomani P, Ulmgren P. A method for decreasing the risk of calcium oxalate scaling in bleach plants: mill trials and flow sheet simulations. 6th International conference on new available technologies, Stockholm, Sweden, 1-4 June 1999, pp 194-199 [Stockholm, Sweden: SPCI Swedish Association of Pulp and Paper Engineers, 1999, 640pp] Richardson B, Lownertz P, Forget C, Uloth V, Gleadow P, Hogikyan R (1998) Behaviour of Non-Process Elements in the Kraft Recovery System Proceedings of International Chemical Recovery Conference, Tappi 1998, p.1025-1039 Åsa Sivard, Tomas Ericsson, Processintegration av vattenreningsteknik – minskning av energi- och resursanvändning i skogsindustrin, Swedish Energy Agency, Project 32218-1, Stockholm, Sweden 2011. Åsa Sivard, Tomas Ericsson, Processintegration av vattenreningsteknik – minskning av energi- och resursanvändning i skogsindustrin, Åforsk, ref 11-179, Stockholm, Sweden 2013. http://www.filtsep.com/view/32856/wood-pulp-and-paper-water-reuse-driveschinese-paper-mill/ (2014-06-02) Per-Åke Frank, Anders Åsblad, Thore Berntsson, Möjligheter att utnyttja värmepumpning i massa- och pappersindustrin – En kartläggning baserad på pinchanalys 625, Värmeforsk, February 1998. Karin Glader, Measures for increased energy efficiency at Iggesund mill, Chalmers university of technology, Gothenburg, Sweden 2011 Anna Fritzson, Simulation and pinch Analysis of Värö, Chalmers university of technology, Gothenburg, Sweden 2002 86 VÄRMEFORSK [17] Henrik Lindsten, FRAM 48, Möjligheter till energibesparingar på Södra Cell Värö med inriktning på indunstningen, Gothenburg, Sweden 2005 [18] Markus Lindahl, FRAM 47 – Design of pulp mill hot and warm water systems – A new method that maximizes excess heat [19] Daniel Ingman, Maria Gustafsson, Mats Westermark, Energieffektivisering inom skogsindustrin genom spillvärmeåtervinning från våtluft – Förprojektering och lönsamhetsbedömning av anläggningsalternativ 1022. Värmeforsk, December 2007. [20] Svenland Torsten, personal communication, Värmeforsk representative Nymölla pulp mill, Värmeforsk reference group meeting, 2014-05-05. [21] Watermark M, Vidlund Anna, Öppen absorbtionsvärmepump för uppgradering av spillvärme från skogsindustrin 955, Värmeforsk, January 2006 [22] Hampus Bellander, Mats Westermark, Verifiering av hygroskopisk kondensering hos Holmen paper Braviken 1200, Värmeforsk, December 2011. [23] A. Teder, U. Andersson, K. Littecke, P. Ulmgren, The recycling of acidic bleach plant effluents and their effect on preparing white liquor, Tappi Journal 73(2), 1990. [24] Paperbref, Deliverable D16: Technical guidelines (water management concepts) for paper makers in European regions with difficult boundary conditions on how to operate mills with minimum water use, Centre Technique du Papier, EC “Energy, Environment and Sustainable Development” Programme 1998-2002. [25] C. Rampotas, H . Terelius, K. Jansson, NETFLOC System – The Tool to Remove Extractives and NPE, Proceedings of 1996 Minimum Effluent Mills Symposium, TAPPI, p. 199. [26] Dubé, M., McLean, R., MacLatchy, D., Savage, P. Reverse osmosis treatment : Effects on effluent quality. Final effluent quality improves. Pulp & Paper Canada 101:8, 2000, pp. 42-45 [27] Ecocyclic Pulp Mill – “KAM”, Final report 1996-2002, Chapter 5: Separation and Purge Systems for Non-Process Substances, KAM report A100, 2003. [28] Åsa Sivard, Thérèse Johansson, Sverker Danielsson, Anders Uhlin, Effektivt utnyttjande av avloppsströmmar I massa- och pappersindustrin. Praktiska försök med utvalda membran och delströmmar, Energimyndigheten (projekt 32218-2), 2014. [29] H. Lindberg, H. Engdahl, R. Puumalinen, strategies for metal removal control in closed cycle mills, Proceedings of 1994 International Pulp Bleaching Conference, Montreal, CPPA 1994. [30] V. J. Böhmer, Recovery of water and chemicals from pulp bleach effluents, Master thesis, water Utilization, university of Pretoria 1992. [31] J. R. Caron, L. D. Williams, Design and integration of the bleach filtrate recycle process, Proceedings of 1996 Tappi Minimum Effluent Symposium, Atlanta, GA, Tappi Press 1996. [32] Magdalena Svanström, Morgan Fröling, Margareta Lundin, Mattias Olofsson, Environmental assessment of supercritical water oxidation and other sewage 87 VÄRMEFORSK [33] [34] [35] [36] [37] [38] [39] sludge handling options, Waste Management & Research, 23 ( 4 ) s. 356-366, 2005. Irma Karat, Advanced Oxidation Process for Removal of COD from Pulp and Paper Mill Effluents: A Technical, Economical and Environmental Evaluation, Master of Science thesis, Royal Institute of Technology, Stockholm, 2013. P. Axegård, J. Carey, J. Folke, P. Gleadow, J. Gullichsen, D. Pryke, D. W. Reeve, B. Swan, V. Uloth, Minimum-impact mills: Issues and Challenges, Proceedings of 1997 Tappi Minimum Effluent Symposium, Atlanta, GA, Tappi Press 1997. Young J.J, Quick J.J; Process element balances in the world’s most closed kraft mill, 65th Appita Annual Conference and Exhibition, Rotorua, April 2011 Association, American Forest & Paper, 2012 AF&PA Sustainability Report P.S. Bryant, E.W. Malcolm, and C.P. Woitkovich, Pulp and Paper Mill Water Use in North America, 1996 TAPPI Environmental Conference Exhibit, Orlando, Florida, May 6-7, 1996. Report on water Conservation in Pulp & Paper Industry, Central Pulp & Paper Research Institute Saharanpur, (UP), India, July, 2008.) Johan Nygaard, Optimal värmeåtervinning I värmeväxlare, Stockholm, Sweden 1977. VDI Heat Atlas, Springer, Second Edition, 2010 88 VÄRMEFORSK Appendices A Earlier research and trends A.1 Earlier research A.1.1 Historical trends in mill closure and pollution prevention Historically, a pulp mill or an integrated pulp and paper mill were counted among industry’s largest fresh water consumers and most serious water polluters. The greatest concerns regarded a discharge of effluents with a high content of organic material, characterized by high biological and chemical oxygen demand (BOD and COD) which lead to serious disruptions of the aquatic life in the recipient. Another factor of great concern included emissions of chlorinated organic compounds, commonly named AOX (adsorbable organic halogens), characterized by high acute toxicity and/or high bioaccumulation. Many of the abovementioned compounds are also classified as carcinogens and endocrine disruptors. Their main source in the pulp mill has been bleaching with the use of chlorine or chlorine-based chemicals. Finally, even other physicochemical properties of the pulp and paper mill have been regarded as problematic from the environmental point of view: high content of dissolved inorganic compounds, high temperature (leading to a thermal pollution) or colour, which affected the quality of the recipient waterways. These environmental concerns where the main factors that lead to the development of various bleach plant closure techniques, and later, to the development of modern effluent treatment facilities. As the older mills during 1970, ’80 and the beginning of ’90 did not have any external effluent treatment facilities, decreasing the effluent volume and its environmental load by internal cleaning and/or recirculation was their only option for decreasing the total environmental impact of the effluents. A report by National Council for Air and Stream Improvement [4] offers an interesting description of the industrial trends in process closure as well as a summary of the developed technical solutions and mill case studies. In short, the mills aimed at gradually adopting a holistic approach to waterborne pollution minimization by: prevention of the generation of large effluent volumes with undesired compounds, reuse of the internal water resources and recovery of the organic compounds from water streams. The above could be accomplished by a combination of several approaches: - Decrease in woodyard water consumption. By replacing some of the wet processes by their dry analogues (e.g. debarking) as well as reuse of some of the less contaminated mills streams (e.g. white water) in woodyard, both water consumption and effluent volume can be decreased. - Improvement in pulping technologies. By adopting novel solutions in pulping processes, e.g. black liquor impregnation, optimization of cooking temperature profile, optimization of liquor dosing and the alkali profile of the cook etc. it became possible to increase the cooking 1 VÄRMEFORSK efficiency and decrease the amount of pollutants generated per ADt of the final pulp products. - Oxygen delignification. Introduction of this technology marked an enormous improvement in both pulping process efficiency and minimizing the environmental load of the effluents. Oxygen delignification is more selective than extended cooking delignification, leading to better utilization of the wood raw material. Moreover, the washing liquors from oxygen delignification are counter currently directed to the recovery plant instead of ending up in effluent streams. All this allowed the mills to both decrease the pollutant generation per ADt of final product and to better distribute the bleaching load, leading to decrease in bleach plant effluents as the kappa number of pulp going to the bleach plant is lower. Nowadays oxygen delignification is accepted as a standard process stage in most market pulp mills. - Optimized washing in cooking, brownstock and bleach plant. Improvements in both washing efficiency (e.g. by application of effective displacement washing and dewatering equipment) and the use of counter current washing allowed minimizing the volume of the effluents as well as the recovery of dissolved organic substances, valuable for energy production in the recovery boiler. - Improved spill management. Large amounts of effluents were traditionally generated in a mill due to spillage, wash losses, extensive wash and rinse routines, overflows etc. For any mills, improving the management of spill and auxiliary water uses in the fiberline, e.g. by identifying the unnecessary water losses and increasing the filtrate storage tank capacities, proved to be both easy and highly effective. - Partial or total elimination of chlorine and chlorine-based bleaching chemicals. Replacement of chlorine bleaching lines with sequences using chlorine dioxide proved to be enormously successful, as ClO2 is much more selective bleaching chemical than chlorine. ClO2 bleaching also generates significantly less AOX than the traditional chlorine bleaching, which lead to significant reductions in AOX emissions from the mills that retrofitted its bleach plants. o Totally chlorine free (TCF) bleaching. The application of modern peroxide and ozone bleaching strategies allowed a total elimination of chlorine-based bleaching chemicals. The advantage includes zero AOX emission and virtually Cl-free bleach plant filtrates. It should be noted that for most mills who invested in TCF bleaching lines, the main motivation was precisely a possibility of obtaining Cl-free filtrates which could be recycled, and not the reduction of AOX emissions, although many mills used and still use the TCF 2 VÄRMEFORSK technology as a sales argument. Södra Cell and Aspa mill were one of the first kraft market actors to develop and introduce fully operable TCF bleaching techniques for highgrade market pulps, based on hydrogen peroxide, peracetic acid and (somewhat later) ozone. Standard chemicals for TCF bleaching are nowadays hydrogen peroxide and a complexing agent (usually EDTA or DTPA) for transition metals removal. - Efficient reuse of evaporation condensates. Weak black liquor from the digester has to be significantly concentrated before combustion in the recovery boiler (from ~15% to over 70-80%). As the typical flow of black liquor solids in a kraft mill is of the order of 1.5 t DS/ADt, it is easy to estimate that substantial amounts of water (~8 m3/ADt) can be available in a form of evaporation condensate. However, the condensate stream is contaminated with reduced sulphur compounds and volatile organic compounds of high COD value. Introduction of a thorough separation of condensate fractions with different contamination levels as well as condensate stripping allowed many mills to reuse different fractions of condensate in various process stages. Clean condensate is nowadays commonly applied in brownstock washing, while the dirtier fractions can be used in the recovery cycle, e.g. for lime mud wash or smelt dissolving. However, a more exhaustive use of condensate for pulp washing may often be restricted by limitations regarding the quality of the final product (e.g. completely odour-free pulp for food packaging production). - Recirculation of bleach plant filtrates (bleach plant closure). During the last few decades, mills have significantly changed their policy regarding liquor flow in the bleach plant. A traditional one was totally open, with fresh water wash almost wherever applicable. Gradually a counter current reuse of filtrate streams within the bleach plant was applied, most often in accordance with desired pH levels. However, a complete purge of the final alkaline and acidic filtrates was still practiced. During the nineties, some mills started to reuse alkaline filtrate as wash liquor in brownstock washing, and in many Swedish mills it is today a common practice. Development of the ECF and TCF bleaching technologies was an important step towards the bleach plant closure. One of the main hindrances in recirculation of bleach plant filtrates was a potential increase of chloride content in the recovery cycle. High levels of Cl- increase the risk for equipment corrosion and decrease the melting point of recovery boiler ESP dust, leading to a troublesome “sticky dust” phenomenon. Switching to technical solutions allowing lower levels of Cl in the filtrates was, for most mills, a necessary prerequisite for bleach plant closure. Recirculation of acidic and Q-stage filtrates proved to be much more difficult than alkaline. An acidic bleaching stage facilitates the removal of most non- 3 VÄRMEFORSK process elements (NPEs) from the pulp and dissolution of the previously formed precipitates. The NPEs become fully soluble and their concentration in acidic filtrates is thus normally much higher than in alkaline ones. Recirculation of such filtrates to brownstock washing may lead to dangerous increase in NPE concentration, scale formation, increase in oxygen-based bleaching chemicals demand and decreased quality of the final product. Some trials have been made with recirculating these filtrates to process stages outside the fiberline (e.g. smelt dissolver) [23] - Advanced purification techniques for the recirculated streams. As stated above, the recirculation of internal streams with high content of dissolved or suspended organics and non-process elements brings about a variety of potential process problems, depending on the type and concentration of the impurities. Recirculation of bleach plant filtrates or other streams with high content of COD-generating organic compounds within the fiberline leads normally to an increased carryover of organics in the pulp stream, which in turn causes larger consumption of bleaching chemicals, needed for chemical decomposition of the extra organics. Moreover, certain organics, e.g. odour-generating reduced sulphur compounds in condensates or wood extractives in filtrates may severely deteriorate the quality of the final product. Recirculation and reuse of the streams with high content of inorganic compounds leads to yet another type of problems. Compounds with high solubility, e.g. K+ and Cl-, may accumulate in the mill, especially in the recovery cycle, leading to corrosion or disturbances in the recovery boiler operation (sticky dust). When it comes to the sparingly soluble elements, like Ca, Mg, Mn, Ba, Si or Al, their accumulation due to filtrate recirculation may lead to serious problems with precipitation and scale formation on the process equipment. Precipitates of CaCO3, BaSO4, Mn(O)(OH) in the fiberline and aluminosilicates in evaporation are common examples. Transition metals, like Mn or Fe, may also significantly increase the consumption of oxygen-based bleaching chemicals and lead to lower pulp quality. Introduction of P and Si to the green liquor system and white liquor preparation leads to lowering of free CaO and increased deadload in the lime cycle. A variety of techniques for removal of both organic and inorganic ions from the recirculated streams has been developed; see the detailed description below. It can generally be said that the higher the separation degree, the higher the investment and/or running costs of the chosen technique. The costs for new equipment, extra energy costs (e.g. thermal for evaporation or mechanical for membrane separation) as well as possibilities for handling the concentrated contaminant stream must carefully be weighed against the gains from lowering the waterborne emissions and obtain an extra water stream. 4 VÄRMEFORSK A.1.2 Purification techniques for internal stream recirculation Below, a variety of purification techniques for internal mill streams are discussed with respect to following classes of contaminants: suspended solids (fibre, large organic fractions), COD-generating organic substances (high-molecular-weight, e.g. lignin till low-molecular-weight, e.g. methanol), easily soluble inorganic ions (Na, K, Cl and certain other inorganic anions) and scale-generating inorganics (Ca, Mn, Mg, Ba, Fe, Zn, Si, Al, P). Flocculation/precipitation. In this process, a clarifying agent is usually added in order to facilitate adhesion of colloidal particles in a suspension, whereby an easily separable solid phase is formed and can be removed from the system [24]. This technique is a standard step in raw water treatment plants or effluent treatment plants, but has been also applied to internal pulp streams, e.g. white water. The most common flocculants include aluminium and iron compounds (e.g. aluminium sulphate, sodium aluminate, iron chloride) and compounds from a polyacrylamide class [24]. Flocculation effectively removes most of suspended solids, certain types of organic substances (e.g. some suspended wood extractives) and possibly some inorganic colloidal precipitates. It is relatively efficient in decreasing the turbidity and colour of the effluent stream. However, as a technique for cleaning internal mill streams for reuse, it is usually not good enough. It also requires costly chemicals and generates some problems with the disposal of separated sludge. A process based on polyethylene oxide flocculants and flotation separation, NetFloc, was developed by Kemira chemicals Oy in cooperation with Domsjö Fabriker MoDo [25]. It was first implemented in Domsjö for sulphite filtrate treatment, but seems to be equally effective on kraft filtrates. It is said to remove even certain problematic metals such as Fe and Ca; according to the provider, removal efficiencies of more than 90% for certain NPEs have been documented. Extractives are removed by reacting with PEO forming a pitch sludge, which can be separated by flotation or screening. After application of NetFloc, Domsjö mill was the first totally effluent free chemical pulp mill in the world [25]. The technique has also reported to be used in Södra Cell Värö mill [4], but is not in operation today. Although it was successfully applied on portion of Q filtrate, the final evaluation showed it was not economically feasible in comparison to external biological treatment, which was finally installed. Biological treatment. Usually applied for the treatment of effluents from the mill before their release. Microorganisms can easily neutralize most of the organic substances giving rise to high COD values, depending on their size and properties (complex macromolecules, like lignin, may require special conditions for effective degradation). Some inorganic compounds, e.g. P or certain metals, are also removed by binding into the sludge formed. The aerobic treatment with activated sludge, aerated lagoon or bioreactor is a common method applied in waste water treatment facilities of pulp mills. Anaerobic treatment has an advantage of much lower sludge production as well as a possibility to recover methane gas formed as a valuable energy source, but it is 5 VÄRMEFORSK not widely applied: in a recent internal investigation by Innventia, only about 10% Swedish mills used an anaerobic treatment stage. However, all the above mentioned techniques normally require rather high COD level in incoming stream, large volumes and residence times, neutralization of the strongly acidic or alkaline streams and stable process conditions. Another disadvantage includes the disposal of the sludge generated, which, according to the newest environmental regulations, cannot be landfilled as long as it is combustible. Biological treatment has therefore not been widely used for internal stream purification. Certain novel technical solutions, e.g. membrane bioreactors, may be considered. Membrane separation. Various membrane techniques have been tested for the removal of unwanted compounds from the mill streams. Membrane filtration is usually conducted in a cross-flow mode, i.e. with the flow of the liquor to be filtered tangential to the membrane, which minimizes the risk for deposition (fouling) of the membrane surface. The net flow of permeate is accomplished due to the pressure difference. Depending on the pore size, membranes can be classified as microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and reverse osmosis (RO) membranes, with decreasing pore size. The lower the pore size, the higher the removal efficiency and the operating costs, as higher pressure differences need to be applied for the same flux (amount of permeate obtained per unit of membrane area and time). For practical purposes, membranes with increased pore sizes are therefore used in series, e.g. microor ultrafiltration as a pre-treatment step for reverse osmosis. While microfiltration is applied for removal of residual suspended solids and some microorganisms, ultrafiltration is more effective in removing both suspended solids, bacteria and in some applications even high-molecular weight organic compounds. COD removal is not very effective and inorganic ions (dissolved non-process element) removal is virtually nonexistent. Nanofiltration can be used to remove COD and bivalent ions such as sulphate and calcium, which is often applied for hardness removal. Monovalent ions, such as sodium or chloride, can only be effectively removed by reverse osmosis. Dubé et al. (2000) [26] reported application of reverse osmosis for the treatment of clean (COD of 1067 mg/L) condensate. 88% of COD and BOD were concentrated in 1% volume as retentate flow. Many trials have successfully been conducted with ultra- and nanofiltration of neutral and alkaline bleaching filtrates [27]. At first, some fouling with high COD liquors was observed. Retention for divalent cations (Ca, Mg, Mn, Ba) was reported to be high, 7080%, probably because they were complexed by organic molecules. The removal of monovalent ions was however poor. COD removal varied depending on the type and origin of the organics (filtrate type, ECF or TCF) and varied between 40-70%. Trials with ultrafiltration of pressurized P stage filtrates showed low COD separation, 20-30% [27]. Divalent metal ions were removed with the efficiency of 40-50%. Nanofiltration doubled COD separation and increased the removal of to 80%. It was concluded that for a best effect a combination of UF/NF would be required, which in turn would imply high operating costs. 6 VÄRMEFORSK Acidic filtrates have been found to cause fouling and degradation of the traditional membranes and required therefore more expensive polyether sulfone membrane material [27]. Trials with membrane purification of Q filtrate showed high flux with ultrafiltration, while nanofiltration was ineffective alone and required pretreatment with larger-pore membranes. COD retention was about 80%, 80% for monovalent and ~100% for divalent ions; i.e. very clean permeate was obtained. Another already mentioned application area of membrane separation included membrane bioreactors. Trials with treatment of wastewater filtrate from PO stage gave 87% COD separation; no data on inorganics removal was reported [27]. Even some recent trials conducted by Innventia and ÅF in cooperation with several Swedish mills [28] suggest that membrane technique can be an attractive option for purification of various mill streams, including filtrates and condensates. Several process parameters need to be optimized: the membrane material, adapted to the physicochemical properties of the liquor, pressure differences necessary for desired fluxes and resulting energy requirements, washing procedures to avoid extensive fouling, pore size etc. In many cases, cascading membranes with increasing pore size gives excellent results, however at the expense of increased investment and operating costs. Chemical precipitation of NPEs. Removal of troublesome non-process elements, like Ca or Mn, by precipitation as carbonates or hydroxides at pH 10.0-10.5 and separation of the solids formed was also proposed. The precipitation would be achieved by the addition of sodium hydroxide and sodium carbonate, gaseous CO2 from flue gases or green liquor [29]. The technique has been tested on acidic and Q filtrates [27] with mixed results. The presence of organic substances with metal-complexing abilities, as well as regular complexing agents, like EDTA, increased supersaturation, hindered the formation of solids and possibly altered their crystalline and morphologic structure, leading often to precipitates which were difficult to separate. This technique may also be inapplicable to removing P, Al and Si, as these NPEs do not precipitate easily at high pH levels. Removal of NPEs by precipitation has therefore never found a large-scale practical application, and is not expected to do so in the nearest future. Evaporation. Evaporation is extremely effective in removing all kinds of non-volatile pollutants, both organic and inorganic. Practically, only water and some volatile compounds (e.g. methanol) can be found in treated liquor. However, high investment and operational costs has severely limited its application. Three main approaches to evaporation have been proposed: traditional boiling-point evaporation [24], low-temperature vacuum evaporation [4] and spontaneous concentration by evaporation in cooling towers [30]. The first of the three approaches has an obvious disadvantage or requiring significant amounts of thermal energy, while the last one involves areas for cooling towers, large flows and problems with microorganism growth. Low-temperature vacuum evaporation by applying mechanical vapour recompression (MVR) technique has been regarded as the most attractive. The technique is said to be suitable for DS 0.5-2% and reduce scaling tendencies compared 7 VÄRMEFORSK to multi-stage evaporation. Capital costs are large, especially if a separate boiler for concentrate has to be built, whereas operational costs are comparable or higher than conventional two-stage activated sludge treatment with tertiary clarification [4]. Evaporation was tested, among others, at Stora Enso Fors mill in Sweden [4]. The primary process effluent from CTMP was evaporated at 7 effects Zedivap (AndritzAhstrom) evaporator from 0.9 to 30-50% DS and sent to nearby kraft recovery for destruction, while condensate was reused. Several other mills, e.g. Stora Enso Kotka and Varkaus mills, Finland, have pilot tested the technology [4], but nowadays it is not reported to be in permanent use in pulp and paper industry. In general, evaporation techniques for internal purification of mill streams are estimated as extremely effective in removing pollutants, but involve high capital and operating costs. Certain problems, e.g. evaporator fouling have also been reported [4]. It is recommended for the mills having possibilities for reusing other equipment for this purpose (e.g. converting old evaporators), access to low-grade heat or cheap electrical energy and whenever high-purity liquid, impossible to achieve by other techniques, is required. Freeze-crystallization. This technique is widely used in food industry for concentrating various products sensitive to high temperatures, e.g. juices. It is said to be more energy efficient than evaporation, as latent heat of freezing lower than latent heat of evaporation. Among other advantages, lower scaling and corrosion tendency, lower capital outlay, lower operating costs and lower VOC emissions were reported. Ice crystals could be separated, melted and water reused. Initial installation at LCTMP Tembec (former Louisiana Pacific) Chetwynd Mill in USA [4] proved unsuccessful due to poor ice crystal growth. Due to problems with poor ice separation, scaling and corrosion in evaporators for the concentrate, freeze crystallization was replaced by vapour-recompression system; neither of these techniques is now in use as the mill was idled in September 2012. No safe recommendations could therefore be given. Ion exchange. This process is commonly used for purification, separation, and decontamination of aqueous and other ion-containing solutions. It utilizes polymeric or mineral material, an ion exchange medium, for selectively exchanging a group of ions for one specific ion until the exchanging capacity is depleted and the medium needs to be regenerated. Usually, separate materials for exchanging cations and anions are used. This technology was applied, among others, in Canton Mill, Evergreen Packaging as a part of custom-developed and patented BRF® (Bleach Filtrate Recycle) technology [4] [31]. The mill was forced to decrease its effluent load due to problems related to complying with colour restrictions, as it released its effluents to a local river located in the area of high natural and recreational value. A fraction of D1 filtrate was filtered to remove fibres and fed to cation exchange which was then periodically regenerated with NaCl. This technique provides no COD removal. Ultimately, 80% closure of D1 and EOP was periodically achieved. According to the mill staff (Innventia 2013) the technique works well for hardness NPEs, i.e. alkaline earth metals, but the removal of transition metals has been found unsatisfactory. 8 VÄRMEFORSK This technique can be recommended for mills having problems with the accumulation of certain non-process elements, without the need of effectively separate organic substances from the purified stream. The high investment and operational costs, as well as problems with the disposal of the spent regeneration liquor need to be considered. Wet-air and supercritical water oxidation. Both these techniques are based on a principle of oxidising organic compounds in the recirculated liquor stream by exposure to gaseous oxygen or air at elevated temperatures and pressures [4]. In supercritical oxidation process, water above its critical point is used as a reaction medium. The residual products are CO2, water and in case of wet-air oxidation, also low-molecular weight organic compounds, like acetic or formic acid. Inorganic compounds are oxidised and precipitated as oxides and carbonates. Their separation is especially effective in supercritical water oxidation technique, as supercritical water is a very good solvent for organic compounds but very poor for inorganics [32]. These two techniques have never been reported to be applied directly to purification of mill filtrates. Experiences from research projects in other fields suggest that for an effective and energy-efficient oxidation, the concentration of organics should be much higher than in a typical filtrate. Oxidation is therefore recommended mostly for sludge, possibly for high-COD municipal waste waters [32]. Purification of bleach plant filtrates would require their concentration prior to oxidation, by e.g. evaporation or membrane processes. Ozone treatment. This technique belongs to a group of advanced oxidation processes, defined as near ambient temperature and pressure water treatment processes which involve the generation of hydroxyl radicals (OH*) in sufficient quantity to affect water purification [33]. Ozone treatment has been studied for treatment of final mill effluents as a part of an external treatment plant. It has been found to effectively reduce colour and COD of the effluent and increase its biodegradability; no reduction in concentration nitrogen and phosphorus compounds was observed. The technique has been estimated as expensive (high costs for ozone generation) and although the COD of the effluent decreased, a biological stage was still required for complete treatment. As an option for internal purification of mill streams it does not seem very attractive, especially taking into account that no reduction in NPE content is achieved. 9 VÄRMEFORSK Table 12.1: Removal efficiencies with various technologies given for five classes of pollutants. After Paperbref [24]. Technology/pollutant Suspended solids Flocculation/precipitation ●● Biological treatment Ultrafiltration ●●● Nanofiltration ●●● Reverse osmosis ●●● Electrodialysis Chemical precipitation Ion exchange Ozone treatment Evaporation ●●● Freeze-crystallization ●●● Wet-air/supercritical ●●● water oxidation COD ● ●●● ● ●●● ●●● ●●● ●●● ●●● ●●● Dissolved inorganics Bacteria Colour ● ● ●● ●●● ●●● ●● ●●-●●● ●●● ●●● ●●● ●● ●●● ●●● ● ●●● ●●● ●●● ●●● ●●● ●●● ●●● ●●● ●●● ●●● A.2 Conclusions from earlier mill closure trials With the development of the environmental science, including new tools for studying the environmental performance of the industrial sites as well as the overall environmental impact of various products, the attitude towards the mill closure has changed. A holistic approach, including life cycle assessment techniques, allowed studying the combined impact of mill emissions at different levels. More focus was put on the fact that every attempt to minimize the effluent load necessarily implied certain material and/or energy expenses. These had to be motivated before any minimization program was put into work. The main conclusion from Alliance for Environmental Technologies panel 1997 [34] stated that a total elimination of bleaching effluents should not be given highest priority. An extensive, or even full, bleach plant closure may even be undesirable if the consumption of raw materials, energy or other resources increases significantly. From 1988 to about 2000, the use of water and effluent flow in a typical pulp mill decreased considerably, and although the raw effluent contaminant levels did so only slightly, the total contaminant discharge from the treated effluents decreased much more. This was caused probably more by improvements in effluent treatment processes rather than the actual closure, although some process modifications were also involved. A noticeable exception was a reduction in AOX emission, accomplished mostly by change in bleaching strategies. Further closure would certainly bring about even more profound reduction in effluent load; it would however lead to an increased consumption of energy and chemicals. Only if particularly strict emission limits or water savings are to be met is a high degree of closure recommended. 10 VÄRMEFORSK For the above reasons, the interest in minimizing the flow of effluents from the mill by increased closure and internal purification techniques has gradually declined during the end of 1990’ and the first decade of 21st century. An external effluent treatment plant has started to be an obligatory part of many industrial sites, including pulp mills. The mills have realised that in most cases when the decrease in the environmental load of the effluents was required, an expansion of, and technical improvements within, the external treatment plant was an easier and more economically feasible option than the mill closure. The closure ideas and numerous internal purification techniques developed at those times were very interesting, but most of them never went beyond pilot tests and mill trials. They constitute nevertheless an invaluable body of theoretical knowledge and industrial experience. A.3 Future trends in mill closure Recently an interest in greater mill closure with an application of some internal purification techniques has started to increase again. Although the current degree of closure, implying an extensive reuse of condensate and commonly practiced recirculation of alkaline filtrate, is usually assessed as sufficient, the mills are becoming interested in theoretical studies as well as simulations of a potential for further closure. There are three main driving forces behind this phenomenon. First, many mills operate within ever tightening economic margins and increase in production is seen as an easy possibility to increase the revenue without a costly retrofit. However, the production increase is difficult to achieve within the current emission limits. It is also expected that the future emission permits may become stricter. Second, increasingly larger amounts of pulp and paper goods are being produced in localisations with limited access to fresh water, or even temporary or permanent water shortages. In certain situations, a mill may even be forced to employ a reduced production mode or even a temporary shutdown. As for now, these problems apply mostly to mills located in South America, but it is expected that in a future even some Scandinavian mills may be affected. A progressing climate change may lead to more weather extremes, alternating heavy rainfalls with lengthy periods of drought. The competition for water resources between industry, agriculture and increasing population of large cities is also expected to increase. Finally, there is a growing interest among customers to buy sustainable pulp and paper products with a high environmental profile. One of the important parameters when evaluating the product’s environmental impact is fresh water consumption during the production stage, often expressed as water footprint. Depending on the exact definition, it may encompass only the direct use of water in the production, but also more indirect effects on water resources, including costs of complete restoration of the quality of an affected aquifer. With the growing environmental awareness among the consumers, it is expected that the mills will become interested in decreasing their products’ water footprint in order to use it as a marketing argument, even if no direct process needs for water savings are present. 11 VÄRMEFORSK It is therefore understandable that when a possibility for water savings or reuse of an effluent stream starts to imply pronounced economic advantages; even some otherwise expensive or complicated purification solutions will become attractive. It is expected that the mills will closely follow any studies on closure strategies. Many Swedish pulp producers begin already to show interest in re-examining the previous studies related to mill closure and internal purification. As mentioned above, the acceptable level of closure is always dependent on the quality of the final product. The goal of water savings or decrease in environmental load has to be therefore weighed against the risk of carryover of unwanted substances into the bleached pulp, which would decrease its value or even render it unsuitable for certain applications, e.g. foodstuff packaging or hygiene products. This is the main reason many mills are reluctant towards full recovery of evaporation condensates, which may, especially during periods of process disturbances, contain malodorous reduced sulphur compounds. A proper control system, allowing a uniform purity of the recovered liquor and shutting down its flow whenever deviations are detected, is necessary. Such systems have to be developed in parallel with closure strategies. There is also a growing interest in a broader closure, encompassing raw water treatment and external effluent treatment. Many mills practice combustion of biosludge from the effluent treatment plant in the recovery boiler, which directs all NPEs separated in sludge to the recovery cycle, where they had to be taken care of. Some mills have reported interest in recirculation of the treated effluent, with an additional purification stage if needed, into the raw water treatment plant or directly into the mill, e.g. fiberline washing. Application of recovered streams in woodyard has also been considered. It is therefore believed that re-evaluation of previous results regarding the available strategies for, and the effects of, mill closure is extremely valuable A.4 References of closed market pulp mills It has been proven to be difficult to fully close the bleach market pulp mill and presently there is no effluent free mill in operation. Thunder Bay was the first pulp mill with extensive closure and full counter current washing in the bleach plant. The mill operated partially closed for many years but was 1978 subject of a recovery boiler tube failure because of chloride induced corrosion [4]. Bleach plant filtrate recovery was resumed in the beginning of 1979 and the reported process effluent was 16 m3/ADt which at that time was extremely low [4]. Ultimately it was decided to terminate recovery of bleaching filtrates and install a secondary treatment facility [4]. The vision of the zero effluent mills became during the 1990-ies again an important subject for many Swedish mills in the work of minimizing the environmental impact. Husum mill was during the 1990-ies operating with extensive recovery of alkaline and acidic bleach plant filtrates to the recovery cycle. The hardwood line ran 25% of the time, completely countercurrently, with no filtrate sent to sewer with averages 5 m3/t effluent on a monthly basis. The softwood line ran with process effluent flow of 8 12 VÄRMEFORSK m3/ADt. Operation was with both TCF and ECF bleaching. The bleach plant was later opened up after construction of new secondary treatment. Other Swedish mills that practised recovery of bleach filtrates to different degrees during the 1990-ies were Aspa, Östrand, Skoghall, Värö and Mörrum [4]. Presently all of the Swedish market pulp mills recover part of the bleach plant alkaline filtrate to the recovery cycle. None of the mills recover acidic filtrate. A.5 References of closed kraftliner mills The kraftliner mill has many similarities with the market pulp mill but one aspect that in this context makes a major difference is that only part or none of the pulp is bleached. These mills recovering alkaline filtrate have TCF bleaching. Kraftliner mills producing white top liner bleach 20-40% of the pulp in a separate production line. This makes it easier to close the bleach plant and recover the alkaline filtrates to the recovery cycle. There are no kraftliner mills in Sweden producing 100% unbleached pulp. The common praxis is to operate these mills with counter-current washing from the paper machine until the oxygen delignification. The white water system in the paper machine is separated from the liquor cycle. There is one example of an extremely closed 100% unbleached kraftliner mill, Visy Paper mill. A.5.1 Visy Pulp and Paper, Tumut, Australia The Visy Pulp and Paper mill, in Tumut, Australia produces unbleached kraftliner board from sawmill logs and wood residues, plus waste paper. Approximately 70% of the finished product is sold to international markets, and the remaining product is supplied to Visy Board where it is processed into cardboard for the packaging industry. Stage 1 of the greenfield mill became operational in May 2001. The design capacity was 300 000 tpy. Since then the mill has been expanded and now has a capacity of 700 000 tpy. In 2013 (1 July 2012 – 30 June 2013) the mill produced approximately 632 000 tons of paper an average fresh water consumption of 3.23 m3/t of paper (monthly averages ranging from 2.85 to 3.62 m3/t. Total waste water production in 2013 was 540 000 m3 (0,85 m3/t) of which 0,23 m3/t was returned to the process and used instead of freshwater. The remaining 0,62 m3/t was discharged to the wastewater re-use scheme. There were no reported discharges of waste water to the recipient Sandy Creek. Water balance The water balance is based on careful design of process water loops to maximise the reuse of condensates and to minimise the use of freshwater to makeup water only. All 13 VÄRMEFORSK paper machine effluent is reused in the fibreline for pulp washing, countercurrent to the flow of pulp. All process water is collected and processed through the evaporators to recover pulping chemicals and to avoid the flow of sulphur and other process materials to other parts of the water circuit. The overall water balance for the mill is summarized in table below. Figures are given as a range to indicate seasonal variations. 14 VÄRMEFORSK Table 12.2: Water balance Visy Paper Parameter Flow [m3/ADt] Water inputs Freshwater Incoming wood Woodyard runoff Rainwater bunded areas Groundwater bores 2.6 – 2.8 1.20 – 1.30 0 0 0 Water outputs Kraft paper Paper machine exhaust Cooling towers Stacks Wastewater Other losses 0.07-0.09 1.1 – 1.3 0.9-0.95 1.0-1.2 0.35-0.40 0.4-0.5 Also a complete NPE balance has been published for the Visy mill [35]. Water inputs Freshwater, including bore water, is separated into a general mill water system and a boiler water system. The mill water system is used for makeup to the WETSAC (turbine condenser), cooling water and seal water systems, and to the paper machines. The boiler feedwater system is composed of micro-filtration, reverse osmosis and electro deionisation. The water in the wood chips ends up in the weak black liquor from the digester, and subsequently in the evaporation plant condensates. The condensates are separated into two streams, intermediate and clean condensate. The intermediate condensate has a higher concentration of organic sulphides, and is used in the recausticising area for lime mud washing, dilution and lime mud filter showers. The clean condensates are used in open areas of the mill such as the paper machine showers, dilution streams and the cooling towers. There is an excess of clean condensate which is discharged to the waste water treatment plant. Surface run-off from the wood yard Area is not suitable for discharge to Sandy Creek due to the potential for it to contain floating wood debris and tannins extracted from wood storage piles. All run-off from the wood yard is collected separately in open drains and channels and directed to the wood yard run-off dam. .A coarse screen is located on the inlet to the dam. Debris captured by the screen is burned in the power boiler. The outlet from the dam is gravity fed to the cooling ponds at the waste water treatment plant. Flow from the dam is regulated to control temperature in the cooling ponds. 15 VÄRMEFORSK All main process areas at the mill site are separately bunded to contain process spills. Each area has tanks and pressure vessels of various sizes and a floor drainage system that transfers spills to a common sump. Each bunded area has been sized to contain 110% of the largest tank. During rain events, runoff enters the mill drains in the bunded areas. The runoff may contain residual chemicals from minor spills, flushing water etc and so the runoff is recovered to the process. Waste water Waste water from the process to the waste water treatment plant is mainly comprised of:     Cooling tower bleed Excess clean condensate from the evaporation plant Boiler feedwater treatment reverse osmosis bleed WETSAC (steam turbine perating) bleed There is a minor flow from the recovery boiler ash treatment system. Other minor losses include water loss in the lime kiln, and grits and dregs. Woodyard run off and domestic sewage are also directed to the waste water treatment plant. Waste Water handling The waste water treatment plant is comprised of two cooling ponds, a sequencing batch reactor (SBR) and 6 ML and 2ML storage ponds. After treatment the effluent has a BOD less than 30 mg/L and total suspended solids less than 50 mg/L. Treated wastewater is re-used for irrigation of the 110 Ha Visy farm property. Irrigation is typically undertaken from October through to May each year. There is a 480 ML winter storage dam for non-irrigation periods. Treated wastewater is also returned to the process as cooling water. In this case the effluent from the water treatment plant is disinfected by chemical dosing A.6 Cooling and secondary heat systems from a historic perspective The specific primary heat consumption in the pulp and paper industry has been reduced significantly during the latest 40 years although the total energy consumption has increased in absolute terms. This can be observed by studying the specific consumptions reported within the energy surveys that Swedish Forest Industry Federation (Skogsindustrierna) has performed since the beginning of the 1970-ies. 16 VÄRMEFORSK Figure 17: Development of energy indicators for Swedish pulp and paper industry. The successively energy improvements in the industry has reduced the primary heat consumption, but also at the same time the amount of available high temperature secondary heat. This simply as a result of the lower primary heat consumption. Typically in the 1970-ies before the oil crisis there was a much lower incitement to save energy in the process. Primarily because of low energy prices, specifically in the secondary heat systems because of the heat surplus caused by high primary heat consumption. Reducing the primary heat consumption and related cost or potential revenues is a major driving force to not only focus on the primary energy, but also the secondary heat. 16 of pulp and paper mills was in a survey performed 2007 asked at which temperature level upgraded secondary heat would be useful, i.e. substituting primary heat or increase the delivery external deliveries district heating [19]. Many of the integrated mills indicated that heat below 80°C would be useful. Also quite many indicated that the temperature level of 80-90°C could be valuable for reducing the primary heat consumption. 17 VÄRMEFORSK Figure 18: Outcome of survey of temperature requirement of 16 pulp and paper mills The author of the article commented specifically today’s lower temperature requirement of secondary heat [19]. This confirms that the temperature requirement has successively gone down with lower primer heat consumption and more efficient processes. The increasing integration of pulp and paper mills with the district heating systems do also reduce the secondary heat surplus and the temperature requirement. 18 VÄRMEFORSK B Reported energy and water consumption data B.1 Water consumptions in an international perspective Water consumptions and resulting effluent volumes have been gathered by EIPPCB and are presented in the BAT BREF 2013 DRAFT report [1] [2]. The BREF-BAT emissions and consumption numbers represents European agreement of the best-availabletechnology within the pulp and paper industry. The above guidelines are proposed in the latest draft dated July 2013, which also is foreseen to be accepted during 2014. Regarding kraft pulp mills the data is collected from EIPPCB questionnaires from 2007/2009 and updated figures from 2010. The reported waste water flow is in the range of 20-90 m3/ADt from bleached kraft pulp mills. The variation is shown in Figure 19 below taken from the BAT BREF 2013 DRAFT report [1]. It is not clearly stated that the waste water flow only includes the process effluent, i.e. no cooling water. When studying the reported effluent flows it can at least be concluded that primarily process effluent is shown for most of the Swedish mills, at least as reported by the mills. Figure 19: Specific waste water flows from mills producing bleached kraft pulp 19 VÄRMEFORSK Note that the Vallvik mill effluent flow includes a large fraction of cooling water consumed at the site. The effluent flow was to a large extent split into process effluent and cooling water effluent when the secondary treatment was commissioned in 2010. Some cooling water effluent is still forwarded to secondary treatment. We also know that some other Swedish bleached kraft pulp mills forward part of their cooling water to the effluent treatment plant, and hence include a portion in the reported process effluent flow. Also note that the actual Värö mill process effluent flow is higher than indicated in Figure 19, as a rather large fraction of the process effluent is by-passed the secondary treatment and instead mixed with the cooling water flow. Both integrated and non-integrated bleached kraft pulp mills are presented in Figure 19. The water consumption per ton of produced pulp is normally higher for an integrated mill compared to a non-integrated mill, one obvious cause being that some integrated mills purchase pulp and produce more paper than pulp. Another cause is that the paper mill also consumes water and generates process effluent and generally more than the drying machine in a non-integrated pulp mill. Just to mention the integrated Swedish kraft pulp mills included in Figure 19 are Karlsborg, Korsnäs, Skärblacka, Gruvön, Husum, Skoghall and Iggesund. In the BAT BREF 2013 DRAFT report [1] waste water (effluent) flows are also presented for a number of unbleached kraft pulp mills, Figure 20. The flows vary between 14 and 80 m3/ADt. Most unbleached kraft pulp mills are integrated with paper mills. It is unclear if the specific waste water flows are expressed per ton of produced unbleached kraft pulp or per ton of produced paper product. If a mill e.g. produces liner, then recycled fibres are usually also included in the paper furnish. Looking into the numbers in Figure 20 it seems as both ways of expressing the specific effluent flows occur. 20 VÄRMEFORSK Figure 20: Specific waste water flows from mills producing unbleached kraft pulp As regarding bleached kraft pulp mills it is not clear if the waste water flow only includes the process effluent, i.e. no cooling water. When studying the reported effluent flows it can at least be concluded that primarily process effluent is shown for most of the Swedish mills, one exception being the Munksund mill. Most surely the cooling water flow is included in the waste water flow. Note that the Frövi and Munksund mills produce significant quantities of bleached kraft pulp in addition to unbleached kraft pulp. The approximate split is 30-40/70-60 % bleached/unbleached kraft pulp, but varies over the years. Corresponding information has been gathered for mechanical and chemimechanical (CTMP) pulp and paper mills that responded to the EIPPCB survey (Source: EIPPCB questionnaires from reference years 2006 and 2008 + Swedish Statistics 2008 etc) [1]. 21 VÄRMEFORSK Figure 21: Specific waste water discharge from individual mechanical and chemimechanical pulp and paper mills Most of the mechanical pulp mills (GW, RMP and TMP) are integrated mills, and their water and energy management is closely linked to the paper mill that manufactures different coated or uncoated paper grades. CTMP mills may be stand-alone mills or be integrated. In mechanical and CTMP pulp mills the water system is fairly closed in order to maintain high process temperatures. In the BAT BREF 2013 DRAFT report they conclude that integrated mechanical pulp and paper mills (GW, RMP and TMP) discharge 9.4-20 m3/ton of product (with the exception of one mill discharging up to 25 m3/t), Figure 21. CTMP mills reported water flows between 9 and 27.2 m3/ADt. With respect to the Swedish integrated mechanical pulp and paper mills the reported waste water flows do not seem to include cooling water flows and most likely this is valid for the other integrated mechanical pulp and paper mills as well. In the BAT BREF 2013 DRAFT report so called BAT-associated performance levels for waste water flows have been defined for the different pulp and paper grades [1]. The performance levels are not emissions levels, but serve as a guide for reducing fresh water use and generation of waste water. The BAT-associated performance levels for the waste water flow at the point of discharge after waste water treatment as yearly averages are [1]: 22 VÄRMEFORSK Figure 22: BAT guidelines for effluent volumes The waste water flows can be regarded as process effluent flows. Note that many paper grades have their origin in two or several of the pulp qualities listed, why it is not obvious how to derive the “correct” waste water performance level for each specific mill and quality. 23 VÄRMEFORSK The American Forest Industry Federation (US) has presented information about the average mill treated effluent discharge from the mid 1970-ies until 2010 (Association u.d.) [36]. The specific effluent flow discharged has been reduced from about 90 m3/ton of product to about 40 m3/ton of product, Figure 23. The effluent data is the average for all pulp and paper mills in the United States and does not include the cooling water effluent that is by-passed the effluent treatment plant. Figure 23: Pulp and paper mill effluent discharges from 1975 to 2010 in the US P.S. Bryant, E.W. Malcolm, and C.P. Woitkovich [36] also reports that the total water consumption for 32 American bleached market pulp mills was 85 m3/ADt in the beginning of 1990-ies which can be compared to the Swedish average of 108 m3/ADt at that time. Regarding the Swedish mills the water consumption can be regarded as the total effluent volume, the process effluent flow plus the cooling water flow. It is although unclear if the specific water consumption = specific effluent for the American mills equals the total flow including cooling water. The specific water consumption for pulp and paper products in Canadian and American mills does not differ significantly. Within the group of developed countries, Spain and Australia have the lowest water consumption of approximately 30 m3 per produced pulp and paper. This can partly be explained by a higher ratio of paper products requiring less water usage per unit, but mainly due to less natural water resources and therefore introduction more technology for water reuse and conservation. 24 VÄRMEFORSK The water consumption many developing countries are significantly higher than in the Scandinavian and North American countries. The water consumption in Indian kraft pulp mills was reported to be 110-220 m3/ADt year 2008. Newsprint mills was in the range 90-160 m3/ton paper [37]. 25 VÄRMEFORSK B.2 Water consumption in Swedish pulp and paper industry Statistics regarding effluent flows, and corresponding raw water consumptions, from the Swedish pulp and paper industry has been collected from Skogsindustrierna. The mills have reported their process effluent flows, cooling water effluent flows etc. as annual figures to Skogsindustrierna. The information is available electronically on the website from year 2001 to 2012. The specific numbers shown in the tables in this chapter have been calculated based on annually produced pulp and paper quantities reported by the mills. Historic data is available in reports issued by NV (Swedish EPA) not published on the website. B.2.1 Swedish market pulp mills In this context market pulp mills refer to mills producing primarily bleached (ECF and TCF) paper grade kraft pulp for the open market. These mills are not integrated with paper mills, i.e. all pulp is sold. The total effluent volume, and corresponding raw water consumption, for Swedish market pulp mills is in the wide range of 30-110 m3/ADt [2]. The process effluent volume (excluding cooling water) is significantly lower. Nowadays process cooling water consumption accounts for a large part of the raw water consumption. Mill A, is the only market pulp mill that has cooling towers and hence does not have any cooling water effluent flow. The process water consumption and process effluent flow is to a large extent dependent of the degree of closure of the fiber line and especially the bleach plant. Swedish market pulp mills generally have a high degree of closure compared to international standards and the corresponding process effluent flow is usually in the range of 30-35 m3/ADt. For example many Swedish market pulp mills, mainly mills with TCF bleaching but also mills with ECF bleaching, recycle and recover part of the bleach plant filtrates. Recycle and recovery of alkaline filtrate is more common than recycle and recovery of acidic filtrate. The BAT-associated environmental performance levels for the waste water flow at the point of discharge after waste water treatment is 25-50 m3/ADt as a yearly average for mills producing bleach kraft pulp, BAT BREF 2013 DRAFT [1]. The waste water flow can be regarded as the process effluent flow. A completely new (greenfield) hardwood market kraft pulp mill is generally designed with a total process effluent flow of 20-25 m3/ADt depending on the specific conditions related to the permit and availability of fresh water at the mill site. For greenfield market kraft pulp mills processing both hardwood and softwood or only softwood the designed process effluent flow is generally 25-30 m3/ADt. Normally there is no cooling water effluent because the mills are designed with cooling towers. 26 VÄRMEFORSK Table 12.3: Specific process- and cooling water effluent flows from different Swedish market pulp mills producing primarily bleached kraft pulp (ECF or TCF) year 2012. Process effluent Cooling water effluent Total effluent TOC COD Nitrogen Phosphorus m3/ADt A 29 C23 17 B 30 D24 69 E25 31 F 34 G 35 7 Mills 35 42 m3/ADt 53 69 24 74 40 33 77 m3/ADt 29 83 86 93 105 74 68 5.4 kg/ADt 2.9 8.0 6.1 7.4 3.5 5.8 4.6 15.6 kg/ADt 7.9 21.5 16.3 22.5 13.0 15.6 12.4 kg/ADt 0.077 0.41 0.29 0.22 0.29 0.21 0.17 0.24 kg/ADt 0.005 0.055 0.019 0.057 0.030 0.036 0.032 0.033 Source: Skogsindustrierna 2012, except for mills F and G which are from 2011, data reported by the mills as annual productions and discharges. Specific discharges have been calculated from annual figures. All mills in table above recycle and recover part of their bleach plant filtrates. For the time being mill A, C and E TCF bleach all pulp, mill B both ECF and TCF bleach their pulp and mill D, F and G ECF bleach their bleached pulp qualities. There may be incorrect allocation of the submitted specific effluent volumes on process water respectively cooling water depending on available measurements. Furthermore, in some few cases cooling water and process water is improperly mixed before the effluent treatment plant. The very low cooling water effluent reported for mill D is misleading. Part of the cooling effluent stream is sent to the secondary effluent treatment plant together with the process effluent stream increasing the process effluent flow. Also the process effluent for mill C is deceptive because it does mainly include the process effluent from the bleaching plant and drying machine area. A fairly large part of the process effluent is bypassed secondary treatment and mixed with the cooling water flow. Further, it should be noted that focus in many bleached kraft pulp mills has rather been to decrease discharges of organic substance (COD, TOC) and other substances rather than extremely decreasing water consumption and process effluent flows. The main reasons are that in Sweden fresh water supplies are good in most areas and that mills try to avoid scaling and operational problems connected to very low water consumption. Historical data of effluent volumes for the Swedish market pulp mills has been gathered and is presented below [5]. 23 Only part of the process effluent is treated. Cooling water flow includes some process effluents. Approximately 20 % is unbleached sulfate pulp. 25 Approximately 20 % is mechanical pulp (CTMP). 24 27 VÄRMEFORSK Table 12.4: Effluent flows (1990 & 2000) from Swedish market pulp mills and average effluent flow for 15 Swedish mills mainly producing bleached kraft pulp. 1990 Process effluent Cooling water effluent Total effluent 2000 Process effluent Cooling water effluent Total effluent F G 15 Mills 130 120 130 108 150 130 120 130 108 28 82 44 42 33 58 NA 28+? NA 82+? NA 44+? NA 42+? NA 33+? NA 58+? Mill A B C D m3/ADt 52 85 130 150 m3/ADt m3/ADt 52 85 130 m3/ADt 36 40 m3/ADt m3/ADt 3627 NA 40+? 26 E NA = Not available Source: NV (Swedish EPA) reports 3925 and 5154 based on data reported by mills to Skogsindustrierna. In year 1990 the mills were asked to report their fresh water consumption, which more or less correspond to their total effluent flow. There is no distinction between process effluent flows and cooling water effluent flows. By 2000 the mills were asked to report their process effluent flows, but not their cooling water effluent flows. Mill A was equipped with cooling towers 2000. However, it is unclear if they had cooling towers by 1990. Like in 2012 the process effluent flow for mill D may be misleading also in 2000. It is probable that part of the cooling water flow is combined with the process effluent flow. From 1990 to 2011/2012 the total effluent flow has decreased significantly in market pulp mills producing bleached kraft pulp, from a range of 50-150 m3/ADt to a range of 30-110 m3/ADt. From 2000 to 2011/2012 the process effluent flow has decreased by about 10 m3/ADt in these mills, from 35-45 to 30-35 m3/ADt. 26 27 Including mechanical pulp production Has a cooling tower 28 VÄRMEFORSK B.2.2 Swedish kraftliner mills The total effluent volume, and corresponding raw water consumption, for the three Swedish kraftliner mills is within quite a wide range, 65-80 m3/ADt when expressed per ton of produced kraft pulp, and 40-55 m3/t when expressed per ton of produced liner, table below. The process effluent volume (excluding cooling water) is significantly lower. Expressed per ton of produced kraft pulp the range is 25-45 m3/ADt, while expressed per ton of produced liner the range is 15-30 m3/t. All three mills use various amounts of recycled fiber in their liner products, 20-45 %. We judge that Swedish kraftliner mills have process water consumptions and process effluent flows in line with international standards. There is no BAT-associated performance level indicated for liner mills in the BAT BREF 2013 DRAFT document [1]. However, environmental performance levels for the waste water flow at the point of discharge after waste water treatment are indicated for unbleached kraft pulp mills and recycled fibre (RCF) paper mills without deinking. The level is 15-40 m3/ADt for unbleached kraft pulp mills and 1.5-10 m3/t for RCF paper mills without deinking. The higher end of the range for the RCF paper mills is mainly associated with boxboard production. Assuming the BAT-associated performance level to be 15-45 m3/t for liner mills the Swedish kraftliner mills perform well with process effluent flows of 15-30 m3/t of liner. 29 VÄRMEFORSK Table 12.5: Specific process- and cooling water effluent flows from Swedish kraft pulp mills producing liner. Specific figures expressed per ton produced kraft pulp and paper (liner) year 2012. Process effluent ditto per ton paper Cooling water effluent ditto per ton paper Total effluent ditto per ton paper TOC COD Nitrogen Phosphorus 3 m /ADt m3/t m3/ADt m3/t m3/ADt m3/t kg/ADt kg/ADt kg/ADt kg/ADt A28 44 29 37 25 81 54 6.2 22.4 0.18 0.050 B29 24 14 45 25 69 39 1.2 8.4 0.23 0.053 C30 25 20 39 31 65 51 1.1 4.2 0.13 0.031 Average 31 21 40 27 71 48 2.8 11.7 0.18 0.045 Source: Skogsindustrierna 2012, data reported by the mills as annual productions and discharges. Specific discharges have been calculated from annual figures. Discharges of various substances to receiving waters from the three mills are given in table above and are only expressed per ton of produced kraft pulp. The numbers would decrease by 20-45% if expressed per ton of produced liner. Historical data of effluent volumes for the Swedish kraftliner mills has been gathered and is presented below [5]. 28 Of total kraft pulp production about 30% is TCF bleached pulp. Liner includes about 35% recycled fiber. 29 Liner includes 45% recycled fiber. 30 Of total kraft pulp production about 30% is TCF bleached pulp. Liner includes about 20% recycled fiber. 30 VÄRMEFORSK Table 12.6: Effluent flows (1990 & 2000) from Swedish kraft pulp mills producing liner. Specific figures expressed per ton produced kraft pulp and paper (liner). 1990 Process effluent ditto per ton paper Cooling water effluent ditto per ton paper Total effluent ditto per ton paper 2000 Process effluent ditto per ton paper Cooling water effluent ditto per ton paper Total effluent ditto per ton paper Mill A31 B32 C33 Average m3/ADt m3/t m3/ADt m3/t m3/ADt m3/t 45 33 NA NA ? ? 15 11 NA NA ? ? 32 26 NA NA ? ? 31 23 m3/ADt m3/t m3/ADt m3/t m3/ADt m3/t 57 38 NA NA ? ? 25 14 NA NA ? ? 21 16 NA NA ? ? 34 23 NA = Not available Source: NV (Swedish EPA) reports 3925 and 5154 based on data reported by mills to Skogsindustrierna. In year 1990 the mills were asked to report their fresh water consumptions. In case of the kraftliner mills it seems as they correspond to the process effluent flows. There is no data available for cooling water effluent flows. By 2000 the mills were asked to report their process effluent flows, but not their cooling water effluent flows. All three mills also used recycled fibres for the production of liner, 10-30% 1990 and about 20-45% 2000. From 1990 through 2000 and to 2011/2012 the process effluent flow has been more or less constant for the kraftliner mills, 10-35 m3/t 1990, 15-40 m3/t 2000 and 15-30 m3/t 2012, all figures expressed per ton of produced liner. B.2.3 Swedish newspaper and magazine paper mills The total effluent volume, and corresponding raw water consumption, for the five Swedish newspaper and magazine paper mills is within quite a wide range, 2060 m3/ADt when expressed per ton of produced mechanical pulp plus deinked recycled fibres if applicable, and 20-45 m3/t when expressed per ton of produced paper, table below. 31 Share of recycled fibre in liner, about 30% 1990 and 2000 Share of recycled fibre in liner, about 30% 1990 and about 45% 2000 33 Share of recycled fibre in liner, about 10% 1990 and about 20% 2000 32 31 VÄRMEFORSK The process effluent volume (excluding cooling water) is significantly lower. Expressed per ton of produced pulp, the range is 15-20 m3/ADt, while expressed per ton of produced paper the range is 10-15 m3/t. Note that when specific figures are expressed per ton of pulp, the pulps included are produced TMP pulp or produced TMP- and groundwood pulp and any handled deinked pulp from recycled fibres, but not purchased bleached kraft pulp if applicable. We judge that Swedish newspaper and magazine paper mills have process water consumptions and process effluent flows in line with international standards. Table 12.7: Specific process- and cooling water effluent flows from Swedish newspaper and magazine paper mills based on pulp production (including any handled recycled fibres) and paper production year 2012. Mill Process effluent m3/ADt ditto per ton paper m3/t Cooling water effluent m3/ADt ditto per ton paper m3/t Total effluent m3/ADt ditto per ton paper m3/t TOC kg/ADt COD kg/ADt Nitrogen kg/ADt Phosphorus kg/ADt A 34 16 15 13 12 28 27 0.8 2.5 0.09 0.006 B35 17 15 7 6 24 21 2.3 5.5 0.07 0.0032 C36 20 15 17 13 38 28 3.1 8.3 0.14 0.009 D37 10 10 12 12 22 22 0.7 2.0 0.04 0.005 E38 19 14 39 28 58 43 1.8 5.0 0.13 0.0032 Average 16 14 18 14 34 28 1.7 4.7 0.09 0.005 Source: Skogsindustrierna 2012, data reported by the mills as annual productions and discharges. Specific discharges have been calculated from annual figures. There are no BAT-associated performance levels indicated for integrated newspaper and magazine paper mills in the BAT BREF 2013 DRAFT document [1]. However, environmental performance levels for the waste water flow at the point of discharge after waste water treatment are indicated for mechanical pulp and paper mills and for recycled fibre (RCF) paper mills with deinking. The level is 9-16 m3/t for mechanical pulp and paper mills and 8-15 m3/t for RCF paper mills with deinking. The BAT-associated performance level that can be derived will be dependent on the mix of pulps in the paper produced in every mill. 34 Approximate mix in paper: TMP 72%, RCF deinked 24%, rest fillers? Approximate mix in paper: TMP 77%, Groundwood 11%, purchased Bleached kraft pulp + filler 12% 36 Approximate mix in paper: TMP 76%, purchased Bleached kraft pulp + filler 24% 37 Approximate mix in paper: TMP 33%, Groundwood 17%, RCF deinked 48%, rest fillers? 38 Approximate mix in paper: TMP 66%, Groundwood 8%, purchased Bleached kraft pulp + filler 26% 35 32 VÄRMEFORSK Without considering the contribution of process effluent from the handling of deinked recycled fibre (mills A and D) and from purchased bleached kraft pulp in the paper machines (mills B, C and E), the five mills will still fulfil the BAT associated process effluent flow of 9-16 m3/t for mechanical pulp and paper mills. The process effluent flows are in the range of 10-15 m3 per ton of produced paper in all five mills. While mills A-D are quite alike with respect to process effluent flows and cooling water flows per ton of pulp or paper, mill E stands out with a high cooling water flow indicating a low degree of cooling water recycle in the TMP pulp mill and the paper mill. Discharges of various substances to receiving waters from the five mills are given in table below and are only expressed per ton of produced mechanical pulp plus deinked recycled fibres when applicable. The numbers would decrease by about 10-25% if expressed per ton of produced paper for mills B, C and D with significant shares of purchased bleached kraft pulp in their paper products. Historical data of effluent volumes for the Swedish newspaper and magazine paper mills has been gathered and is presented below [5]. In year 1990 the mills were asked to report their fresh water consumptions. In case of the newspaper and magazine paper mills it seems as they correspond to the process effluent flows. There is no data available for cooling water effluent flows. By 2000 the mills were asked to report their process effluent flows, but not their cooling water effluent flows. The mix of pulps in the paper products have changed over the years and details for all of the five mills are not available for 1990 and 2000. However, from 1990 through 2000 and to 2011/2012 the process effluent flow has decreased for the newspaper and magazine paper mills, 12-25 m3/t 1990, (8)13-18 m3/t 2000 and 10-15 m3/t 2012, all figures expressed per ton of produced paper. Regarding mill D it may be noted that the process effluent flow has jumped up and down during the years presented, though on a low level, 12 m3/t 1990, 8 m3/t 2000 and 10 m3/t 2012. Ever since 1990 the mill would more than well be in line with international standards of 2013 (BAT-associated performance levels 2013). 33 VÄRMEFORSK Table 12.8: Effluent flows (1990 & 2000) from Swedish newspaper and magazine paper paper mills. Specific figures are based on mechanical pulp production (including any handled recycled fibres) and paper production. 1990 Process effluent ditto per ton paper Cooling water effluent ditto per ton paper Total effluent ditto per ton paper 2000 Process effluent ditto per ton paper Cooling water effluent ditto per ton paper Total effluent ditto per ton paper A B C D E Ave. m3/ADt m3/t NA NA 23 21 25 24 13 12 21 19 21 19 m3/ADt m3/t m3/ADt m3/t NA NA NA NA NA NA ? ? NA NA ? ? NA NA ? ? NA NA ? ? m3/ADt m3/t 17 16 14 13 26 18 8.5 8 16 14 m3/ADt m3/t m3/ADt m3/t NA NA ? ? NA NA ? ? NA NA ? ? NA NA ? ? NA NA ? ? 16 14 NA = Not available Source: NV (Swedish EPA) reports 3925 and 5154 based on data reported by mills to Skogsindustrierna. B.3 Water treatment and quality in Norwegian and Swedish mills An investigation regarding Swedish and Norwegian raw water treatment was performed during 2013. The aim of the study was to find out the technical status of the equipment and possible concerns regarding the raw water handling. The study included preparation of process water but not boiler feed water. If there were technical solutions of general interests, this was of course also of interest. The investigation is based on questionnaires and/or interviews and is performed to get a good overview of the situation and not to investigate detailed information about the installations. Totally 34 mills from Norway and Sweden are included in the study. B.3.1 Age of installations In Figure 24 the ages of main equipment in the different raw water installations are presented. 34 VÄRMEFORSK Figure 24: Ages for main equipment in raw water installations in Sweden and Norway There are several probable reasons for quite old installations like:    Not prioritized part of the mill Decreasing water consumption per ton product produced, will result in no need for investments for higher raw water capacity and therefore no new equipment. Quit robust equipment with a long life time. However, there are quit many mills considering exchanging their raw water treatment at the moment to reduce maintenance costs. B.3.2 Chosen process In Figure 25 the type of different processes for main equipment in the installations for raw water treatments are presented. 35 VÄRMEFORSK Figure 25: Type of process solutions for raw water treatments. As can be seen in Figure 25 most installations consist of a combination of mechanical and a chemical process steps. In many cases the mechanical and chemical treated water are used at different positions in the mill where both qualities are produced. Chemical treated water is for example preferred for chemical handling and for production of bleached chemical pulp and mechanical treated water will be used at less sensitive positions. There are also some plants that do not use any kind of treatment. These are mainly Norwegian mills producing mechanical paper with a very good raw water quality and one case is unbleached pulp production. A couple of mills have chosen to use municipal water for emergency. The water will be used when the raw water treatments of their own cannot produce a water of high enough quality due to seasonal variations. Another reason for using the municipal water as emergency is if the river will change direction and salt water from the sea will enter the raw water intake. Munkedal paper mill has chosen to reuse biological treated water in the mill. 50% of the biological treated water will be filtrated in an ultrafiltration plant and after passing a polishing pond sent back to the mill. The main reasons for the installation are low environmental permits by authorities. The discharge levels out from the mill are at the lower level compared to IPPC BAT levels. The installation is shown below: 36 VÄRMEFORSK Figure 26: Recirculation of process water for papermill. The treatment plant consists of 6 process steps a, free swimming bacteria b, biobed c, moving bed biofilm reactor d, sedimentation e, ultrafiltration f, polishing pond with recirculation Also a second Swedish mill producing tissue paper recirculates a part for the treated effluent water to the water intake. 37 VÄRMEFORSK B.3.3 Type of water treatment compared to type of production Figure 27: Type of treatment for different kinds of production Figure 27 shows that the most chemical treatments are installed for bleached sulphate pulp production and fine paper. These productions are also most sensitive for impurities in the water. There are also cases where the raw water quality is good enough for bleached sulphate production and cases where the mechanical paper productions have chosen to use chemically treated water. The clear tendency is however that a cleaner product will use more highly processed raw water. 38 VÄRMEFORSK B.3.4 Preheating of raw water Figure 28: Preheating of raw water before treatment. There can be positive effects from preheating raw water before the treatment process. Some plants have chosen this before entering the raw water treatment, as can be seen in Figure 28. These measures will give a potential to obtain positive effects like:   Better performance and operational costs for the chemical treatment plant Energy optimization by reusing heat from other flows The preheating is obtained by heat exchanging different streams in the mill. Another chosen solution is to combine the warm outlet for cooling water and other warm streams with the intake for process water, which will result in an increased temperature for the raw water. Especially in the winter when the raw water is cold, these measures have a positive impact on the process. 39 VÄRMEFORSK B.4 Heat and power consumption in Swedish mills The heat and power consumption in Swedish mills is briefly discussed within this study to give a perspective on the potential savings related to reducing fresh water consumption and improving secondary heat systems. The energy consumption for all Swedish pulp and paper mills has been gathered from the beginning of the 1970-ies until today by the Swedish Forest Industries Federation [7]. Internationally there is no comparable source available on such a detailed level and for such long time period. Information about the energy usage specifically related to cooling systems and secondary heat systems is however difficult to find for Sweden or in the international literature. Based on available data it is although possible to compare the presently reported specific power and heat consumption for different Swedish mills with the specific water consumption to see if there is any correlation. B.4.1 Market pulp mill The heat and power consumption for the Swedish market pulp mills has been extracted for analysing the development with respect to heat and power consumption. The average specific steam heat consumption and electric power consumption for producing kraft market pulp is presented in table Table 12.9 below. Also the minimum and maximum reported consumption for a single production unit is showed. Sulphate market pulp dried in flash dryers is not included in the collected data. Table 12.9: The development of the specific energy consumption for Swedish market pulp mills from 1973-2011 compared to specific water usage. Market pulp production Steam heat consumption (max) (min) Power generation (max) (min) Power consumption (max) (min) Average process water consumption (excl. cooling water) 2011 kADt/year 3372 2007 3715 2000 3557 1994 2894 1988 2945 1984 2832 1979 2082 1973 2097 GJ/ADt 13.9 14.4 15.2 15.4 15.1 15,1 17.3 16.6 kWh/ADt 15.7 10.9 680 21.2 12.2 643 16.2 11.5 538 17.4 11.0 523 19,8 12,4 507 19,2 12,8 493 24.4 13.8 553 21,4 14,8 412 1035 429 796 927 700 827 330 800 987 681 775 67 798 1019 664 845 185 854 1043 680 709 0 837 1433 708 837 0 841 1073 665 734 230 846 1056 687 784 1140 600 kWh/ADt 35 58 40 108 VÄRMEFORSK It is clear that the steam heat consumption has been reduced considerably since the beginning of the 1970-ies. During the same time period the water consumption has been reduced significantly with higher degree of process closure. It is however difficult to separate the part of the reduced heat consumption that is related to the lower water consumption. Higher degree of closure generally increases the temperature in the process and reduces the need for primary heat usage (steam heating). This typically observed in the process stages after the digesting, i.e. brown-stock washing, bleaching and the wet end on the pulp dryer. These are also the areas where most of the hot water is consumed in the process. Today’s lower water consumption for washing of pulp may partly explain the reduced primary heat consumption; also more energy efficient recovery of secondary heat for generation of more hot water is also an explanation. The steam heat and power consumption for the today’s Swedish market pulp mills is presented below. There is no clear correlation when comparing the specific energy consumption with the specific water consumption. Table 12.10: Specific energy consumption for Swedish market pulp mills 2011 and the relative water consumption. Steam heat GJ/ADt consumption Power kWh/ADt generation Power kWh/ADt consumption Relative water consumption A 15.7 B 15.5 C 15.2 D 15.4 E 14.8 F 12.7 G 11.9 556 1035 712 429 570 706 815 786 895 760 700 835 771 722 Low High Low Low High Low High The increased process closure and last decade’s expansion of the effluent treatment plants may actually to some extent increase the power consumption. 41 VÄRMEFORSK B.4.2 Kraftliner mills Historic power and steam heat consumption has been extracted for kraftliner mills in Sweden [7]. Table 12.11: Historic specific energy and water consumption for Swedish kraftliner mills. Liner production kton/year 2011 1448 2007 1614 2000 1637 1994 1565 1988 1271 1984 1134 1979 930 1973 583 Steam heat consumption Power consumption Average process water consumption (excl. cooling water)39 GJ/ton 4.5 4.5 4.8 4.6 4.4 4.8 5.9 6.1 kWh/ton m3/ton 483 458 535 518 472 440 464 515 21 23 23 The presented consumption data is related to the paper machine and does not include the energy consumption related to pulp production. The kraftliner sulphate pulp production has a similar development as the market pulp production. The ratio of recycled has increased gradually since the 1970-ies which reduce the overall energy consumption related to pulp preparation. The gradually decreasing the steam heat consumption is most probably related improved technology for pressing the pulp web to higher dryness before entering the drying section of the paper machine. B.4.3 Magazine paper mills Historic power and steam heat consumption has been extracted for magazine paper mills in Sweden [7]. 39 The average water consumption is reported for 2012, 2000 and 1990. The water consumption is expressed per ton kraftliner and includes both the paper and pulp production. 42 Nr VÄRMEFORSK Magazine paper production Steam heat consumption Power consumption Average process water consumption (excl. cooling water)40 kton/year 2011 1133 2007 1158 2000 788 1994 469 1988 354 1984 398 1979 306 1973 347 GJ/ADt 5.1 5.2 6.0 7.1 6.7 7.6 8.3 8.8 kWh/ADt m3/ton 776 804 711 774 712 825 737 786 14 14 19 The presented consumption data is related to the paper machine and does not include the energy consumption related to pulp production, i.e. refiners and grinders. The trend shows a continuous reduction of steam heat used for production of magazine paper. The major part of the reduction is related to higher pulp dryness after the press section. Also the main part of the pulp used for the products was changed from sulphite and grinded pulp to thermo mechanical pulp (TMP) improving the possibility of achieving higher press dryness. During the last decade the amount of LWC paper which can be pressed to high dryness has increased, which also explains the lower reported steam heat consumption. Steam heat for hot water generation in paper machines is generally lower and any possible improvement during the years cannot explain any larger part of the reduced consumption. 40 The average water consumption is reported for 2012, 2000 and 1990. The water consumption is expressed per ton kraftliner and includes both the paper and pulp production. 43 Rubrik Table 12.12: Historic specific energy and water consumption for Swedish magazine paper mills. VÄRMEFORSK C Description Bleached market pulp – Reference mill C.1 Design criteria The design of the kraft market pulp mill represents today´s state-of-art technologies and is based on best available and commercially proven technology. The design of the reference mill considers:  high, consistent pulp quality which is competitive on the international market  the product is elemental chlorine free (ECF)  low specific consumptions of wood, chemicals, and water  high energy efficiency  maximized production of bio-energy, and minimal usage of fossil fuels  low environmental emissions; on the level of newer modern mills  cost-effective solutions The reference mills are not based on equipment from any one supplier. In general the key process data used in the balances in this study are conservative and should not exclude any of the major pulp mill equipment suppliers C.1.1 Production and mass balance The softwood raw material consists of 50% pine (Pinus sylvestris) and 50% spruce (Picea abies). The relation between roundwood with bark and sawmill chips is 70% roundwood and 30% sawmill chips. The softwood debarking is performed in dry debarking drums which are designed for a barking efficiency of 95%. The bark is used as fuel in the lime kiln and power boiler. The digester is a continuous digester with white and black liquor impregnation. The cooking temperature is 143oC and the Kappa number after the cooking is 30. Oxygen delignification is done in two stages without intermediate washing to a kappa number of 12. Oxidized white liquor is the primary alkali source. The brown stock washing consists of three wash presses prior to the oxygen stage and two wash presses after. The brownstock washing dilution factor is 2.5 m3/ADt. The carryover of COD from the oxygen delignification to the bleach plant is calculated to be perati 5 kg COD/ADt, excluding the bleach plant filtrate re-circulated to brown stock washing. 44 VÄRMEFORSK Figure 29: One example of a typical brownstock washing system (Metso). The softwood is bleached to a final brightness of 90% ISO. The bleach plant is designed with four bleaching stages. The first stage is operated as a “conventional” D-stage, and the sequence is D(EOP)DP. Wash presses are used for all washing in the bleach plant. The bleach plant is designed to release 10-15 t/ADt of effluent. This range includes an allowance for up to 5 t/ADt of fresh water. Figure 30 shows the liquor flows in the bleach plant. White water from the pulp machine is used as wash liquor on the wash press after the P-stage. The filtrate from this wash press is then used as wash liquor on the 2nd D stage wash press. Fresh water is used as wash liquor on the (EOP) stage wash press and condensate is used as wash liquor on the 1st D-stage press. The filtrate from the (EOP) wash press is then transferred as wash liquor to the 2nd wash press after the oxygen stage. Hot / warm water Clean condensate Hot water Chemicals 4.1 t/ADt 4.1 t/ADt ~2 t/ADt ~ 5 t/ADt HD O2 D D EOP P To 1st O2 From dryer washer 4.5 t/ADt To treatment ~10 t/ADt To treatment ~5 t/ADt effluent effluent Figure 30: The approximate liquor flows (t/ADt) of the ECF bleach plant. The dilution factor is about 2 t/ADt. The pulp dryer is of floating web type, which dries the pulp web while keeping it floating on a cushion of hot air heated by LP (3.5 bar(g)) steam. Exhaust air from the pulp dryer goes through the heat recovery system which includes heat exchangers for preheating of drying air and air for building ventilation. The evaporation plant is a conventional 7-effect system utilizing LP and MP steam. It is designed to produce 80% dry solids liquor (including recovery boiler ash). All 45 VÄRMEFORSK evaporator bodies are of the falling-film type, and the seven effects are designed to operate in counter-current fashion, i.e. with live steam being fed to the first effect and weak liquor to the seventh effect. The body operating with the strongest liquor in the first effect is heated by intermediate pressure steam from a steam ejector driven by MP steam and compressing LP steam. The other two bodies in the first effect are heated by LP steam only. In the reference mill the recovery boiler is designed to produce high pressure steam at 100 bar(g) and 505C. The mills are equipped with conventional causticizing with both green liquor and white liquor filtration. The green liquor is filtered in two parallel green liquor filter units. The dregs are washed and dewatered in a filter press before being discharged. Condensate from the evaporation plant is used for dregs washing. Dregs and grits are combined and sent to landfill. In table below key operating data for the reference mill is summarized. For further information of the process layout see [5]. 46 VÄRMEFORSK Figure 31: Summary of the reference mill key operating data. ADt/a Softwood 2 000 355 92% 1 045 000 Wood yard Wood to digester Bark and wood waste t/24 h t/24 h 4072 420 Digester Plant Kappa number Unscreened deknotted digester yield Alkali charge on wood as effective alkali Sulphidity (white liquor) % NaOH,% mole-% 30 47.0 20.0 35 Oxygen Stage Kappa number after oxygen stage Alkali charge as NaOH Oxygen charge kg/ADt kg/ADt Washing Department Dilution factor in the last stage m /ADt unbl. 3 2.5 Evaporation Plant Weak black liquor to evap. (incl. spill) ditto dry solids content Strong black liquor, DS content incl. ash Total evaporation, including spill t/h % % t/h 913 16.0 80 771 Recovery Boiler Estimated higher heating value (virgin) Strong liquor virgin solids to mixing tank Net useful heat from liquor (virgin) Net useful heat from liquor MJ/kg DS t/24 h MJ/kg DS MW Dried pulp from dryer Operating days Mill availability Annual production Causticizing and Lime Kiln Causticizing efficiency Total white liquor production Lime kiln load Active CaO in lime Lime kiln fuel ADt/24 h d/a mole-% 3 m /24 h t/24 h % 47 12 25 20 14.0 3 477 10.3 413 82 7 541 534 90 Bark / wood waste VÄRMEFORSK C.2 Energy balance The designed reference mill is very energy efficient and the black liquor alone produces enough steam to satisfy the process steam consumption in each of the mills. There is also an excess of steam from the recovery and power boilers which is utilized in a condensing turbine to produce in green power which is sold. The lime kiln is fired with bark powder, or gasified bark, and the remaining bark from the woodyard and chip screening is burned in a power boiler. Some key items are summarized as:               HP steam data 100 bar(g), Recovery boiler sootblowing steam is extracted at 25 bar(g) from the turbine instead of using HP steam Feed water preheating to 175oC to increase HP steam generation Recovery boiler flue gas cooler to reduce LP steam consumed in air preheating Top preheating of all recovery boiler combustion air to 205oC Latest technology for pulp digesting which has a lower cooking temperature than other systems 7 effect evaporation plant Digester steam consumption has increased slightly with the new liquor extraction High temperature of hot water, 85oC, and maximum use of hot water instead of steam in the bleach plant, and pulp machine Steam consumption in the bleach plant is reduced; more chlorine dioxide and less hydrogen peroxide allow a lower bleaching temperature compare to the design 2004. Low pressure steam used in the pulp dryer Dryness from the shoe press to the dryer is increased to 53-55% depending on wood specie. Pressurized condensate system Bark press for bark to the power boiler Table below presents the overall steam and power balances for the 2010 reference mill. For more information see [5]. 48 VÄRMEFORSK Table 12.13: Summary of steam and power balances. Reference 2010 SW STEAM BALANCE (GJ/ADt) Generation Recovery boiler Power boiler Secondary heat Total steam generation 17.79 1.60 0.60 19.99 Consumption Process steam Back pressure turbine Condensing turbine Total steam consumption 9.00 3.20 7.78 19.99 POWER BALANCE (kWh/ADt) Generation Back pressure power Condensing power Total power generation 866 788 1654 Consumption Mill consumption Sold power Total power consumption 600 1054 1654 C.3 Mill water distribution and cooling system The reference mill process cooling requirement is met by recirculating cooling water over a set of cooling towers. The raw water is treated in the water treatment plant. The mill water system has only one quality, chemically treated water, with the following treatment sequence:     Water intake with coarse screening. Chemical treatment in a dissolved air flotation (DAF). Sand filtration. Clear water well, including storage capacity for firefighting. As precipitation chemical some kind of Al-salt and polymer is used. Raw water sludge is discharged to the receiving recipient together with treated effluent. The cooling water system is semi-open, which means that part of the process water comes from the cooling water system. There is a separate cooling water loop for the turbine and the turbine condenser. The cold fresh water make-up is blended into the cooling water going to the turbine 49 VÄRMEFORSK condenser. The water from the turbine oil cooler, as well as other oil coolers, is dumped to minimize the risk for contamination of the mill water system. Other coolers in the mill are connected to the general mill process water system. Water from such coolers that could contaminate the water should also be dumped. There are filters in the cooling water system to avoid impurities in the mill process water. The amount of process water coming from the process cooling system is controlled so that the cold water temperature is maintained at about 18°C. Figure 32: Process diagram for water treatment plant and cooling systems. In addition to normal heat losses of different kinds, approximately one third of all the energy that is introduced with the fuel to the system will have to be cooled away by a cooling system. The reference mill process water consumption as well as the secondary heat and cooling system are described in appendix H.1. C.4 Secondary heat system The reference mill secondary heat system start with the warm water (50°C) produced from cooling water in the evaporation plant surface condenser. The main part of the warm water produced in the evaporation plant is returned to the cooling towers. A smaller part is sent to the mill warm water tank. The warm water storage buffer is relatively small because there is always a surplus of warm water coming from the evaporation plant. The warm water is distributed to several coolers where semi-hot water (65°C) is produced. This semi-hot water is used in a second stage for producing hot water (85°C). The high hot water temperature reduces the need of steam heating in-between the bleaching stages. Normally there is a surplus of hot water that is returned to the cooling 50 VÄRMEFORSK towers. This excess hot water is on its way back to the cooling tower used for preheating of boiler-make up water and the turbine condensate. The hot water is stored in a storage tank typically with a size of 1000 m3 for new installations. The major part of the hot water consumers in fibre line and causticising do not return the water. However, some hot water consumers such as boiler feed water preheaters return the chilled hot water to the warm water tank or semi-hot water header. The reference mill does not have any external consumers of hot water such as district heating, biomass dryers. The design of the reference mill secondary heat system would be slightly modified if different external consumers would be connected to the system. 51 VÄRMEFORSK D Description Kraftliner – Reference mill D.1 Design criteria The kraftliner reference mill is designed to produce two qualities of liner (brown and white top) on two parallel paper machines from unbleached liner pulp and recycled fibre pulp produced at the mill, plus purchased bleached kraft pulp. The design of the mill is based on best available and commercially proven technology, as found in European mills. The kraftliner mill considers the same general design principles related product quality, environment and energy to as listed for the market pulp mill (se Appendix C). The reference mill production basis and mass balance is presented. Also the energy balance and main features is discussed. The process concept for the water treatment, cooling water and secondary heat systems is discussed more in detail. Details in other mill areas such as, evaporation plant, cooking, fiberline etc. can be found in the full Reference mill report [5]. D.1.1 Production and mass balance In this study the kraft pulp mill has a maximum continuous rate (MCR) of 2000 ADt/d. The total liner production is 3100 t/d including recycled fibres and purchased bleached pulp. Table below summarizes the key operating and dimensioning data and for the reference mill. 52 VÄRMEFORSK Table 12.14: Summary of pulp mill key operating data. Unbleached pulp Operating days Mill availability Annual pulp production ADt/24 h d/a ADt/a 2 000 355 92% 653 200 Wood yard Wood to digester Bark and wood waste t/24 h t/24 h 3 221 303 Digester Plant Kappa number Unscreened deknotted digester yield Alkali charge on wood as effective alkali Sulphidity (white liquor) % NaOH,% mole-% Washing Department Dilution factor in the last stage m /ADt unbl. 3 2.0 Evaporation Plant Weak black liquor to evaporation, excl.spill ditto dry solids content Strong black liquor, dry solids content incl. ash Total evaporation, including spill t/h % % t/h 636 14.1 80 564 Recovery Boiler Estimated higher heating value of virgin DS Strong liquor virgin solids to mixing tank Net useful heat from liquor, virgin solids Net useful heat from liquor MJ/kg t/24 h MJ/kg DS MW Causticizing and Lime Kiln Causticizing efficiency Total white liquor production Lime kiln load Active CaO in lime Lime kiln fuel 85 56.0 17.0 35 14.4 2 128 10.6 260 mole-% 3 m /24 h t/24 h % 82 4754 336 90 Bark/wood waste Table 12.15: Summary of recycled fibre plant key operating data. RCF production, at kraft mill MCR Annual RCF production ADt/d ADt/a 650 212 000 To match the large capacity of the kraft pulp mill the paper mill has two paper machines: 53 VÄRMEFORSK  PM1 produces 525 000 t/year unbleached kraftliner, with up to 50% recycled fiber in the base ply.  PM2 produces 481 000 t/year white top liner. The corresponding pulp requirements are:  RCF 212 000 ADt/year  Bleached kraft pulp (purchased) 164 000 ADt/year  Unbleached kraft pulp 653 000 ADt/year Table 12.16: Summary of liner machines’ key operating data. Machine speed, at pope Width on pope Grammage Fibre furnish Unbleached kraft pulp Bleached kraft pulp Recycled fibre (RCF) pulp Consumption, at kraft mill MCR Unbleached kraft pulp Bleached kraft pulp RCF pulp Paper production, at kraft mill MCR Annual paper production m/min m 2 g/m % % % ADt/d ADt/d ADt/d t92/d t92/a PM 1 PM 2 1340 7.5 120 (80-200) 1230 7.5 120 (115-200) 67 0 33 59 33 8 1097 0 548 903 502 100 1610 525 000 1473 481 000 D.2 Energy balance The reference mill is designed to be energy efficient. However, due to the relatively high liner production based on recycled fibres and purchased bleached pulp, the mill is not in fuels, and must purchase additional wood fuel for the power boiler. The steam consumption in the mill co-generates power. Even though the paper machines and RCF plant have low power consumption there is still a need to buy power. Some key items in the reference mill include:     HP steam data 100 bar(g), 505oC Feed water preheating to 175oC to increase HP steam generation Recovery boiler flue gas cooler to reduce LP steam consumed in air preheating Top preheating of all recovery boiler combustion air to 205oC 54 VÄRMEFORSK        Latest technology for pulp digesting which has a lower cooking temperature than other systems; the steam consumption is however about the same 7 effect evaporation plant Dryness from the shoe press to the dryer is about 49% for PM1 and 48% for PM2 Recovery boiler sootblowing steam is extracted at 25 bar(g) from the turbine instead of using HP steam Pressurized condensate system High temperature of hot water, 85oC, and maximum use of hot water instead of steam in the bleach plant, and pulp machine Bark press for bark to the power boiler Table below presents the overall steam and power balances for the 2010 reference mill. Table 12.17: Summary of steam and power balance Model 2010 GJ/ADt unbl pulp GJ/t liner STEAM BALANCE Generation Recovery boiler Power boiler Secondary heat Total steam generation Consumption Process steam Back pressure turbine Total steam consumption POWER BALANCE Generation Back pressure power Purchased power Total power generation 11.25 2.56 0.23 14.04 7.26 1.65 0.15 9.06 11.06 2.98 14.04 7.13 1.92 9.06 kWh/ADt unbl pulp kWh/t liner Consumption Kraft mill Paper mill Recycled fiber plant Sold power Total power consumption 799 513 1312 515 331 846 479 775 58 0 1312 309 500 37 0 846 D.3 Mill water distribution and cooling system The reference mill Kraftliner secondary heat and cooling system is rather similar to the market pulp mill. The main difference is related to the water usage on the paper machine and RCF-plant, as well as the non-existing bleaching plant. Also the utilization of the 55 VÄRMEFORSK evaporation plant condensates becomes different. Approximately 2.2 m3/ADt (1000 mg/l COD, 65°C) intermediate condensate (B-condensate) is used in the causticising plant. The remaining condensate is clean which is discharged as effluent (200 mg/l COD, 80°C). The kraftliner mill is also equipped with cooling towers for minimizing the fresh water consumption. There is although no condensing turbine and related cooling water consumption in the kraftliner mill. Principally, the raw water treatment process and cooling system is the same as for the market pulp mill. It is a semi-open system where hot water system has the same water quality as the cooling water circuit, i.e. part of the cooling water from the evaporation plant surface condenser is used for production of hot water. The reference mill process water consumption as well as the secondary heat and cooling system is described in Appendix 0 D.4 Secondary heat system The evaporation plant surface condenser is designed for a warm water temperature of 50°C. The small part of the produced warm water is used for hot water production, but the major part is recirculated to the cooling towers. Hot water is produced in the dissolving tank vent condenser and in the cooking plant. Produced hot water is stored in a hot water tank before it is consumed in the pulp washing, causticising and paper machine. There is an excess of hot water at 90°C that is returned to the cooling towers after preheating of boiler feed water. E Description Magazine Paper – Reference mill The integrated SC-paper mill is designed to produce magazine paper (SC paper) from bleached mechanical pulp. The mill is based on the best available and commercially proven technology in the Nordic countries. The model mill production basis and mass balance is presented. Also the energy balance and main features is discussed. The process concept for the water treatment, cooling water and secondary heat systems is discussed more in detail. Details in other mill areas such as, evaporation plant, cooking, fiberline etc. is can be found in the full Model mill report [5]. E.1 Design criteria E.1.1 Production and mass balance 56 VÄRMEFORSK The paper mill capacity is designed for 456 000 t/a of SC-paper, which means an annual average of 1284 t/d. The paper mill has one paper machine for SC-paper production with average basis weight 56 g/m2. The SC-paper furnish consists of 58% bleached TMP, 12% bleached kraft softwood pulp and 30% filler. The kraft pulp is purchased and slushed in the mill. SC-paper is normally used for magazines and printed in web heatset offset or rotogravure printing. The corresponding TMP capacity is designed for 934 ADt/d, with an annual mean production of 766 ADt/d. The TMP plant is designed to produce a low freeness pulp for SC-paper. The final freeness is 30 CSF. The bleach plant is a two stage process with peroxide and alkaline as the main bleaching chemicals. The target brightness of the TMP is 75% ISO. The process water is taken in a counter current flow from the paper mill via the bleach plant and the TMP plant to the effluent treatment. This is important for the bleach process and the paper machine runnability. Table below summarizes the key operating data for the reference mill. 57 VÄRMEFORSK Table 12.18: Summary of operating data for the reference mill. TMP mill Production, design Yield Availability Operating days per year Production net, annual average Production net Energy input Total motor load ADt/d % % D/a ADt/d ADt/a kWh/ADt KWh/ADt SC-paper machine Speed design Width on pope M/min M Grammage PM furnish composition TMP Chemical pulp Filler Paper mill efficiency Operating days per year Paper production net, annual average Paper production net TMP consumption, annual average Chemical pulp consumption, annual average Filler consumption, annual average G/m 2 934 93 95 355 766 272 000 2700 3050 1850 10.5 56(4860) % % % % D/a t/d t/a ADt/d 58 12 30 82 355 1284 460000 766 ADt/d T100/d 158 356 E.2 Energy balance The model mill is self-sufficient in steam from the TMP process, and there is actually a surplus of steam. Generated steam is used for drying and heating in the paper mill. Bark, and sludge from the effluent treatment plant is sold, as there is no power boiler in the mill. The main features of the reference mill energy concept are:    Dryness from the press section to the dryer is 52% Power consumption in the paper mill is 600 kWh/t paper 77% heat recovery of TMP refiner power consumption Table 12.19 summarizes the overall steam and power balance. 58 VÄRMEFORSK Table 12.19: Summary of steam and power balance STEAM BALANCE Consumption Paper machine Drying Miscellaneous Total steam consumption Model 2010 GJ/ADt GJ/ADt unbl TMP paper 4.63 0.75 5.37 2.87 0.46 3.33 Generation TMP heat recovery Secondary heat Total steam generation 7.63 0.07 7.70 4.73 0.04 4.78 Steam Surplus 2.34 1.45 kWh/ADt unbl TMP kWh/ADt paper 3050 1010 4060 1810 600 2410 - - POWER BALANCE Consumption TMP plant Paper mill Total power consumption Production E.3 Mill water distribution and cooling system The magazine paper reference mill water distribution and cooling system use the same technology as described for the market pulp mill Appendix D. The raw water is mechanically and chemically treated and is distributed to the process. The cooling water system is semi-open. The main part of the fresh water is used on the paper machine for high pressure cleaning showers in the wire- and press sections and for dilution of chemicals. The process related fresh water consumption on the paper machine is about 4.5 m3/ton of paper and in total for the mill about 6.5 m3/ton of paper. The effluent flow is 5.8 m3/ton paper. The reference mill process water consumption as well as the secondary heat and cooling system are described in Appendix 0. E.4 Secondary heat system The fresh water make-up is used as cooling media before it enters the drying section heat recovery system. The produced warm water has a temperature of 50-55°C and is stored in the warm water tank before distributed to the process. 59 VÄRMEFORSK Cooling water is collected separately and is recirculated to the fresh water system and warm water tank. 60 VÄRMEFORSK F Mass balances and block diagrams F.1 Bleached market pulp – Reference mill 61 VÄRMEFORSK F.2 Kraftliner – Reference mill 62 VÄRMEFORSK F.3 Magazine paper – Reference mill Figure below shows the concept for the TMP plant. The main line refining is done at high consistency in three stages. The rejects from the screening and hydro cyclone cleaning plants are dewatered and further refined at high consistency in a two stage reject refining. Refined and screened pulp is dewatered before storage in an MC-storage tower. After the MC-storage the pulp is further dewatered to high consistency before peroxide bleaching in two stages. The first stage is MC (10%) and the second stage is HC (30%). The pulp is then diluted with paper machine white water before entering the paper mill. Wood yard Chip bin Steam Chips washing Pre-steaming Steam Impregnation Chelating agent Pre-heating Refining 3 stages Steam cyclones Steam Electric energy Steam Reboiler Latency chest Electric energy Screening Rejects refininng 2 stages Hydro cyclone cleaning Steam cyclones Disc filter Screening & Cleaning Steam HC storage Bleaching 2 stages Chemicals To paper machine Figure 33: Mass flow diagram for the magazine paper reference mill 63 VÄRMEFORSK A block diagram of the paper machine process is shown in figure below. White water to bleaching plant & slushing of chemical pulp Bleached Kraft TMP Bleached Kraft storage tower TMP chest Refiner Refiner Dosing chest Dosing chest Filler Screen Dosing chest Mixing chest Machine chest Filler Storage tower Cleaner and deaeration Filter Broke tower Screen Screen and deaeration Headbox Wire silo Wire section Pulper White water tank Press Pulper Dryer Pulper Disc filter White water tower Reel Calender Pulper Finishing Figure 34: Block diagram for the magazine paper reference mill 64 VÄRMEFORSK G Energy balances G.1 Bleached market pulp – Reference mill 65 66 VÄRMEFORSK 67 68 VÄRMEFORSK G.2 Kraftliner – Reference mill 69 70 VÄRMEFORSK 71 G.3 Magazine paper – Reference mill The TMP plant produces more steam than the paper machine needs for drying of the paper. The surplus is 1.45 GJ/t paper. Below is a table with steam and power balance of the mill. Table 12.20: Model mill steam balance Clean steam balance GJ/ADt unbleached TMP GJ/t paper 4.63 0.75 5.37 2.87 0.46 3.33 TMP heat recovery Secondary heat Total production 7.63 0.07 7.70 4.73 0.04 4.78 Clean steam surplus 2.34 1.45 GJ/ADt unbleached TMP GJ/t paper 2.69 2.69 1.67 1.67 3.81 2.43 2.42 1.66 10.32 2.36 1.51 1.50 1.03 6.40 7.63 4.73 Consumption Paper machine Drying Miscellaneous Total process consumption Production Contaminated steam balance Consumption TMP plant Total process consumption Production Primary refiners Secondary refiner Tertiary refiner Reject refiners Total production Contaminated steam surplus 72 VÄRMEFORSK Table 12.21: Model mill power balance. kWh/ADt unbleached TMP kWh/t paper Power consumption TMP plant 3050 1810 1110 660 Secondary refiner 670 400 Tertiary refiner 670 400 Reject refiner 600 350 Paper mill 1010 600 Total consumption 4060 2410 Primary refiners 73 H Water balances H.1 Bleached market pulp – Reference mill Figure 35: Seconadry heat and water balance for market pulp reference mill 74 VÄRMEFORSK H.2 Kraftliner – Reference mill Figure 36: Secondary heat and water balance for kraftliner reference mill. 75 Freshwater 7.1 18.0 oC 0.2 0.2 75 oC Make-up water 76 Figure 37: Secondary heat and water balance for magazine paper reference mill 0.6 -1.23 GJ/t92% Flow Temp 20 oC 23 33.0 oC 40oC Cooling in Refiners 0.62 GJ/t92% Cooling tower 30oC 0.70 GJ/t92% 1.0 Miscellaneous Heat recovery system 20 oC General cooling Cooling of Condensate 0.04 GJ/t92% t/92% paper C o 24 33.0oC Wet dryer vent Clean Steam C C 4.2 20.9 oC 51.4 o 0.6 54.4 o 0.3 0.8 Outlet data To external treatment Cooling tower Wet dryer went Paper Process ventilation Consumtion Fresh water Water in chip Water balance 82.1 oC TMP plant Paper machine Wood yard Auxillary 4.7 54.5 oC 5.9 0.6 1.1 0.1 0.0 7.7 t/92% paper 7.1 0.6 7.7 Process ventilation Paper Sum of reject from process 5.9 37 oC Effluent treatment Cooling tower -0.3 GJ/ADt 5.8 48 oC Magazine paper – Reference mill VÄRMEFORSK I Optimal design of heat recovery units The question of how far it is economically feasible to drive the heat recovery is often raised when designing new heat recovery systems or when a new unit is added to an existing unit. The optimal design of heat recovery units was studied as early as 1977 [38] at ÅF. The following chapter and Appendix I presents a general method of determining the optimal temperature difference for sizing of heat exchangers. The main conclusion from the analysis is that heat recovery units should be generously sized considering today’s high energy cost of primary heat in Swedish pulp and paper mills. Typical cost level for heat exchangers results in a theoretic optimal sizing corresponding to a minimum temperature difference of 1-9°C on the hot or cold side. It is important to consider non-ideal counter current conditions (channeling in heat exchangers) when a very low temperature difference can be justified. I.1 Heat exchanger definitions The review of optimal sizing of heat recovery units is based on the following definitions: Figure 38: Heat recovery or cooling unit consisting of a countercurrent heat exchanger. The symbols have the following definition. Equations are generally expressed without any dimension. G c T Q k d n Mass flow Specific heat Temperature Transferred heat Heat exchanger transfer coefficient Operation time Value of recovered heat or cost of cooling media 77 kg/s kJ/kg, °C °C kW kW/m2, °C h/year SEK/MWh F a Total heat exchanger area Desired gross income per marginally installed heat transfer area. Directly related to IRR. m2 SEK/m2, year Figure 39: The temperature profile for the heat exchanger is presented in figure above. Figure 39 assumes that the heat capacitivity for G1 x c1 is equal or higher than G2 x c2. I.2 Determination of optimal temperature difference The optimal temperature difference ΔT depends on the ratio between the heat capacity (kJ/°C) of the cold and hot flow that is heat exchanged. The optimal temperature difference for a given heat exchanger cost is highest when the cold and hot flow have the same heat capacity, i.e. ratio is 1. Normally the heat exchanger design itself limits this relation to be within the range 0.2-5. Tubular heat exchangers can accept a higher range compared to plate heat exchangers. The theoretical optimal design point at maximum continuous rate (MCR) has been plotted for a “base case” representing typical marginal heat exchanger cost and value of primary energy. The base case assumes the following conditions:      Marginal value of recovered heat is 150 SEK/MWh (defines value of primary heat saving). Assumed 8000 hours operation per year (defines value of primary heat saving). Cost for marginal heat exchanger area is 4000 SEK/m2 (defines equipment cost). Heat transfer coefficient is 1200 W/m2,°C (defines equipment cost). Required return on invested capital is equal to 2 years simplified pay-back (defines equipment cost). The optimal design point has also been calculated for two additional cases representing double value of primary heat or half equipment cost, as well as half value of primary heat or double equipment cost. Selecting a design point on any of these two design lines is analogous to doubling or halving any of the above presented design parameters. 78 VÄRMEFORSK The optimal temperature difference (y-axis) is plotted against the logarithmic scale of the heat capacity ratio (x-axis) in the diagram presented below. Figure 40: Design diagram for heat recovery units The analytical solution to the above design chart is presented in appendix I. I.2.1 Example 1 In this example assumed the following conditions:     There is an effluent flow of 6 kg/s with a temperature of 65°C (hot medium). There is cold water with an initial temperature of 5°C and flow of 3 kg/s should be preheated. variable cost related to the heat exchanger sizing is 4000 SEK/m2. heat transfer coefficient is 1.2 kW/m2,°C. operation time is 8000 hours per year. 79  value of primary heat (steam) is 150 SEK/MWh. The conditions are the same as for the “base case” defined above. The ration of heat flow capacities is 6 kg/s x 4.2 kJ/kg,°C devided by 3 kg/s x 4.2 kJ/kg,°C = 2. The black arrow shows how the diagram should be used for finding the optimal design point of the heat recovery unit. The optimal temperature different for the heat recovery unit is estimated to 2.7°C. The temperature of the cold water is after heat exchanging against the effluent is 62.3°C. I.2.2 Example 2 The same condition as in example 1 but the effluent flow is 3 kg/s, the primary heat value is 120 SEK/MWh and the heat exchanger marginal cost is 6400 SEK/m2. The ratio between the “base case” equipment cost and primary heat value and example 2 data is calculated as (150/120)*(6400/4000) = 2. Consequently, the curve representing double equipment cost or half heat value should be used. The ration between the heat capacities based on the new mass flows is calculated as 3/3=1. The green arrow shows how the diagram should be used for finding the optimal design point of the heat exchanger. The optimal temperature different for the heat recovery unit is estimated to 9°C. The temperature of the cold water is after heat exchanging against the effluent is 58°C. I.3 Other practical aspects to consider in the design of heat recovery units Lower temperature difference than about 1°C is difficult to accomplish, this due to effects such as channeling in the heat exchangers. Channeling is caused by different pressure drop over the individual plates/tubes resulting in different volumetric flow. Tubular heat exchanges can also suffer from channeling on the shell side. The effects of conduction and mixing in the direction of flow needs to be considered when heat exchangers are designed for very low temperature differences (< 1°C). The above diagram assumes ideal counter current flow is not valid for very low temperature differences ΔT. A logarithmic mean temperature difference correction factor F then needs to be used. This factor is listed in for example VDI Heat Atlas for different types of heat exchangers [39]. It is difficult to reach temperature difference in the range 1-5°C in real operation because of non-ideal flow patterns. Minimizing channeling effects through heat exchanger design is consequently of importance in situations where a high value of the heat saving justifies a very low temperature difference. 80 VÄRMEFORSK One standard procedure to minimize the channeling effect and improve the heat recovery is to do the heat exchanging in steps (two or three heat exchangers in series) and in units which are long and narrow. This is a common procedure for tubular heat exchangers. Heat exchangers in series means higher pressure drop and needs to be considered in the design. There is also plate type heat exchangers designed with a multiple passage where the fluids are heat exchanged counter currently against in several steps. This gives low channeling that allows very low ΔT. I.4 Analytical solution to heat exchanger optimization problem When solving the equations for the optimal heat exchanger size, the following assisting expressions are formed: Ratio between heat capacity of flows (m >1). Difference in temperature between medias, original temperature difference Minimum temperature difference in heat exchanger. Maximum temperature difference in heat exchanger. Logarithmic mean temperature difference Ratio between decired gross income per installed heat transfer area and income per area at original temperature difference. The numeric value that results in an optimal heat exchanger, i.e. a gross income “a” from the last installed heat area, can be calculated as the following expression: 81 which also can be expressed as A special case comes up if m = 1, i.e. the heat capacity is equal on both sides of the heat exchanger, resulting in the same temperature difference all over the heat exchanger. Another special case is if m = ∞, as if G1 would be condensing steam. In this case the equations become: or The minimum temperature difference Δt1 becomes in this case independent of the original temperature difference. The calculation assumes that the condensation heat in flow G1 is sufficiently high to heat flow G2. When the optimal Δt1 has been calculated, the required heat transfer area can be perating based on the following equations: (Heat transfer area, m2) It is important to note that the above method only gives the optimal size of the heat transfer area, but does not give an answer of the total economy related to heat recovery and cooling. The total economy becomes dependent of how larger the fixed costs related to the installation alone. In some cases with limited space it becomes essential to define a heat exchanger installation cost per area (SEK/m2) in addition to the heat exchanger material cost (SEK/m2). All above calculations are valid if the cooled media has a higher heat capacity compared to the heated media. If the opposite relation would take place, the values needs to be 82 VÄRMEFORSK rearranged so that m is larger than 1. In this case the equations for the minimum temperature difference over the heat exchanger becomes: Figure 41: The temperature profile for the heat exchanger where m<1. The determination of the optimal heat exchanger sizing is based on a given heat capacity on both sides, in other words a fixed “m”. This is a relevant approach for heat recovery units since there is seldom freedom to select the process side flow, and the cold media temperature is given. The flexibility in the cooling media flow (kg/s) is in the higher end rather limited by the available heat quantity in the hot process flow, but also the minimum accepted temperature (example mill hot water temperature) of the media leaving the heat exchanger on the heat exchanger. I.4.1 Example 1 There is an effluent flow of 6 kg/s with a temperature of 65°C that can be used for preheating 3 kg/s cold water with a temperature of 5°C. The design and installation requires a fixed investment cost of 100 000 SEK. In addition to this comes a variable cost of 4000 SEK/m2 related to the heat exchanger. The heat transfer coefficient is 83 1.2 kW/m2,°C and the operation time 8000 hours per year. The heat saved replaces steam with a cost of 150 SEK/MWh. For marginally installed heat exchanger area a capital cost corresponding to 50% IRR is applied (simplified pay-back of 2 years). With a specific heat capacity of water of 4.18 kJ/kg,°C the calculation becomes: The calculations thereby show that it is economically feasible to drive the heat recovery relatively far. As far as 95.6 % of what is theoretically possible to recover with respect to the heat capacity on the cold side of the heat exchanger (1-Δt1/ΔT). Based on the total incoming effluent flow, calculated above the incoming cold water side, 47.8 % heat is recovered. 84 VÄRMEFORSK J Optimal sizing of coolers When replacing or installing new heat exchangers for only cooling purposes it is relevant to review what is a suitable sizing is taking into account both the investment cost, but also the water consumption and related operational cost in the raw water treatment and distribution. The optimal design point of coolers differs significantly compared to heat recovery units A large heat exchanger would have a lower cooling water consumption and operation cost, but the capital cost is high. A small heat exchanger would have high cooling water consumption and power cost for pumping water, but low capital cost. Therefore, the suitable sizing will depend on the power price and power consumption, relative the investment cost for the heat exchanger installation and the pay-back criteria that is applied. Finally, the water conservation itself can be given a value. Such a value is relevant to include if the raw water treatment plant is capacity limited or if the mill operates close to the environmental permit. There is also a marketing value to have low water consumption. This “soft” aspect is not given an economical value in this study. For heat recovery units discussed in chapter 7.3.3, the value of recovered heat can be defined fairly well by calculating the value for marginally saved primary heat. This value is rather independent of the flow because the operational cost related to pressure drop is very low in relation to the primary heat value. A typical cooling water cost has been calculated in Appendix B. The cost has been expressed both in SEK/MWh cooling duty and SEK/m3 water consumption. The latter is a better because the operation cost for cooling water is mainly related to the water volumes that are transported through the system. To determine the suitable sizing of new coolers, the total cost including both capital and operation cost has been calculated for different heat exchanger sizes. The optimal sizing plot has been prepared for cooling of 300 kg/s process effluent from 65°C to 45°C with cooling water of a temperature of 15°C. The thermal load is 25.1 MW. The cooling water distribution pump and raw water pumping station has a total head of 10 m and the pump overall efficiency is set to 80%. The heat transfer coefficient is set to 1200 W/m2,°C The power price is assumed to be 500 SEK/MWh, and the pay-back criteria determining how the capital cost should be related to operational costs, is set to 2 years for new heat exchangers. 85 The total cost (SEK/MWh) on the (y-axis) is plotted against the temperature increase (°C) of the cooling water. A larger heat exchanger allows higher temperature increase and correspondingly lower volumetric flow for a given cooling duty. Figure 42: Numerical solution for the optimal size of an effluent cooler. Plotting the total cost for above preconditions results in an optimal heat exchanger size corresponding to a temperature increase of 13°C. The resulting cooling water flow is 460 kg/s and heat exchanger size is 630 m2. This corresponds to the optimal relation between capital cost for the heat exchanger purchase and the operation cost given the assumed conditions. The recommended design for a new cooler does not necessarily correspond to the economic optimal design calculated above. It can be observed that the total cost increases only marginally (≈10%) when doubling the temperature increase from 13°C to 26°C. However, the size and capacity of the cooler increase by approximately 25% to 790 m2 which gives useful design margins for future higher production. Also the design margins for upset conditions needs to be reviewed when designing a new cooler. Selecting a heat exchanger sized based on the cooling water temperature increase at the economic optimal design at MCR multiplied by 2 is a good rule of thumb that gives appropriate design margins for most installations. However, the capacity needs to be checked for the most demanding operation case, and if the rule of thumb size is not sufficient, the sizing is instead determined by upset conditions, not the operation cost vs. capital cost. A heat transfer calculations shows that the process flow design margin for the theoretical optimal cooler design of 630 m2 corresponds to about 60% compared to MCR operation. This calculation is based on the assumption that the cooling water flow can be doubled temporarily during upset conditions. 86 VÄRMEFORSK A similar calculation for the 790 m2 heat exchanger with 10% higher life cycle cost results in 100 % design margin compared to MCR. The arbitrary design with 40°C temperature increase resulting in an 85 % larger heat exchanger of 1160 m2 compared to optimal design gives a design margin of 300 %. The total life cycle cost increases by 60%. J.1 Determination of optimal temperature increase on cooling water The above example describing the optimal sizing for an effluent cooler can also be applied other cooler installations with differed preconditions. The optimal sizing of new coolers is dependent of the available cooling water temperature and the temperature of the heat source that is cooled. The initial temperature difference between the cooling water and heat source is proportional to the suitable temperature increase of the cooling water (Tout-Tin). The optimal temperature increase has been calculated for a number of temperatures differences between cooling water and the heat source (Tout-Tin). A “base case” has been established assumes the following conditions:       Value of marginally consumed power is 500 SEK/MWh (defines value of primary heat saving). Assumed 8000 hours operation per year (defines value of primary heat saving). The cooling water distribution pump and raw water pumping station has a total head of 10 m and the pump overall efficiency is set to 80% (defines value of primary heat saving). Cost for marginal heat exchanger area is 4000 SEK/m2 (defines equipment cost). Heat transfer coefficient is 1200 W/m2,°C (defines equipment cost). Required return on invested capital equal is 2 years (defines equipment cost). The optimal design point has also been calculated for two additional cases representing double value of primary heat or half equipment cost, as well as half value of primary heat or double equipment cost. Selecting a design point on any of these two design lines is analogous to doubling or halving any of the above presented design parameters. 87 Figure 43: Design chart for coolers The above optimal design points has been determined assuming the heat source is cooled 10°C. Any deviation from this assumption affects the optimal design point only marginally and no correction factor has been developed. J.1.1 Example 1 In the example there is an effluent flow of 300 kg/s that should be cooled from 65°C to 45°C with cooling water of a temperature of 15°C. The cooling water distribution pump and raw water pumping station has a total head of 10 m and the pump overall efficiency is set to 80%. The heat transfer coefficient is set to 1200 W/m2,°C. Power price and payback criteria are set to 500 SE/K/MWh and 2 years, respectively. The conditions are the same as for the “base case” defined above. The initial temperature difference of the heat source and cooling water (x-axis) is 65°C-15°C = 50°C. Reading the above diagram for the base case according to the black arrow results in an optimal temperature increase of the cooling water corresponding to 14°C. The origin of the small deviation from the numerical solution to the problem in Figure 42 is that the design chart is based on a standard 10°C cooling of the heat source. 88 VÄRMEFORSK The rule of thumb sizing is that the optimal temperature of 14°C is doubled to give good capacity margins for upset conditions and future high production. The suitable sizing is 28°C. J.1.2 Example 2 The same condition as in example 1 but the electric power value is 400 SEK/MWh and the heat exchanger marginal cost is 6400 SEK/m2. The ratio between the “base case” equipment cost and primary heat value and example 2 data is calculated as (500/400)*(6400/4000) = 2. Consequently, the curve representing double equipment cost or half heat value should be used. The initial temperature difference of the heat source and cooling water (x-axis) is 65°C15°C = 50°C. Reading the above diagram for the base case according to the black arrow results in an optimal temperature increase of the cooling water corresponding to about 10°C. The rule of thumb sizing is that the optimal temperature of 10°C is doubled to give good capacity margins for upset conditions and future high production. The suitable sizing is 20°C. J.2 Increase size and number of existing coolers The suitable sizing of new coolers or replacing old coolers is discussed in chapter 7.3.4. The optimal sizing is relevant from energy and water consumption point of view. The possibility of increasing the size or number of existing coolers to reduce the water consumption is reviewed in this chapter. Increasing the heat exchanger area makes it possible to increases the cooling water exit temperature and thereby reduces the flow for a given cooling duty. It is although not always beneficial to increase the size of existing coolers. The performance can actually be reduced if the pressure drop and fluid velocity in the heat exchangers is low. J.2.1 Installing an additional heat exchanger The economic feasibility of doubling the size of coolers by installing a second heat exchanger in series or parallel is estimated. The calculations has been done specifically for a coolers operating with a low rise in the cooling water temperature which can be found in existing mills where the production been increased stepwise without improving the capacity in process coolers in the same rate. 89 The typical cooler with low temperature rise is assumed to operate with a inlet cooling water temperature of 15°C and an exit temperature of 25°C. The cooling duty is assumed as 10 MW. The feasibility analysis of adding an additional cooler has been done for both parallel and cascade configuration. Installing the second cooler parallel would reduce the flow rate in the existing heat exchanger (split of flow on two units) with a significant reduction the heat transfer rate on the cooling water side. Installing the second heat exchanger in cascade would maintain a higher flow rate in the existing heat exchanger the overall thermal performance would be higher. This configuration may although in some cases be limited by the pump because of the increased pressure drop. The calculations has also been done for cooling of heat sources at three (3) different temperature levels, 40°C, 50°C and 60°C. Cooling of process medias at higher temperature than 60°C is normally combined with some sort of heat recovery with a higher requirement of thermal performance of the heat exchanger. Cooling of process medias below 40°C is rather uncommon. Principally heat sources at a lower temperature level does not justify as high temperature increase of the cooling water (with resulting lower flow per MW) compared to hot heat sources. Mechanically treated raw water is assumed as cooling media and the cooling water distribution pumps are assumed to operate with a fixed head of 5 m. An extra allowance of 5 m for the pressure drop in the raw water treatment plant is included. The potential water saving and energy saving for the different cases is estimated below for the six (6) different cases. It is important to understand that the process conditions and heat exchanger installation cost needs to be reviewed for each individual cooling position. The below feasibility analysis gives a more general indication of the feasibility of further study of installing additional heat exchangers for reducing the energy and water consumption in Swedish pulp and paper mills. The installation cost is based on a plate type heat exchanger. The total cost includes an estimated allowance for piping, erection and process control. 90 VÄRMEFORSK Table 12.22: Resluts for additional cooler Case 1 Additional cooler Parallel Case 2 Additional cooler Parallel Case 3 Additional cooler Parallel MW m3/h 10 857 10 857 m2 °C 125 15 °C Pressure drop Heat transfer on cooling water sidae Average temperature of heat source Cooling water flow after installation of a second heat exchanger Cooling water outlet temperature [kW/m2] Water saving Power saving OPEX reduction Capital cost Simplified pay-back Cooling duty Original cooling water requirement Heat exchanger area Cooling water inlet temperature Cooling water exit temperature Case 4 Additional cooler Case 5 Additional cooler Case 6 Additional cooler 10 857 Cascade 10 857 Cascade 10 857 Cascade 10 857 83 15 62 15 125 15 83 15 62 15 25 25 25 25 25 25 With new cooler Lower Much Reduced 50 With new cooler Lower Much Reduced 60 With new cooler Higher Reduced With new cooler Higher Reduced With new cooler Higher Reduced °C With new cooler Lower Much Reduced 40 40 50 60 m3/h 714 680 670 463 343 274 °C 27.0 27.6 27.8 33.5 40.0 46.3 m3/h MWh/y kSEK/y kSEK y 143 33 16 325 19,8 177 41 20 249 12,3 188 43 21 212 9,8 394 90 45 325 7,2 514 118 59 249 4,2 583 134 67 212 3,2 A reasonable return on capital invested in a larger cooler capacity is assumed to be equal to 2 years simplified pay-back. It can be concluded that none of the simulated cases can be justified. Installing a completely new cooler can rarely be justified by the power saving. An additional or a completely new cooler can only be justified because of required cooler capacity to satisfy the process requirements. However, when an new cooler is required, it is recommended that the sizing considers the energy and water consumption as outlined in chapter 7.3.4. J.2.2 Adding plates on an existing plate heat exchanger The capital cost for increasing the heat exchanger area by adding plates in an existing heat exchanger should be significantly lower than installing a completely new cooling unit. It is however very difficult to give a general cost estimated of the marginal cost since the it varies significantly between heat exchanger types, materials etc. 91 Adding more plats is analogous to installing a second parallel heat exchanger. Based on the analysis in Appendix K , it is judged that it is difficult to justify adding plants for reducing the cooling water flow. J.2.3 Replacing existing heat exchangers Replacing existing heat exchangers with the purpose of reducing the cooling water consumption is difficult to justify based on the analysis in Appendix K above. J.3 Theory of optimal design of coolers Some typical cooling water cost has been calculated in appendix K. The cost has been expressed both in SEK/MWh cooling duty and SEK/m3 water consumption. The latter is a better because the operation cost for cooling water is mainly related to the water volumes that are transported through the system. The cost per consumed m3 cooling water can be recalculated to cost per removed MWh based on the following relationship: The specific cooling cost is thereby inversely dependent of the how much heat that is transferred per m3 cooling water. Practically the operation cost for pumping the water is lower per MWh for a given m3 if the outlet cooling water is increased. This fact that the selected cooling water flow has a big impact on the cooling process makes it necessary to solve the equations presented in appendix I.4 above for a range of different cooling water flows to find the optimal heat exchanger sizing from operational point of view. It is although not possible to solve the below equation with a cooling water cost (SEK/MWh) which a function of G2, ΔT4 and ΔT2. The minimum temperature difference Δt1 for optimal heat exchanger sizing needs to be determined numerically. 92 VÄRMEFORSK The equation above can be rewritten as: The equation must be solved numerically with respect to Δt1. If the same effluent flow as in example 1 (appendix J.1.1) must be cooled from 65ºC to 45ºC with cooling water of a temperature of 5ºC, what is the optimal heat exchanger size? The cooling water distribution pump and raw water pumping station has a total head of 10 m and the pump overall efficiency is 80%. When solving the equation numerically the optimal heat exchanger efficiency is determined to 45%. This means that the optimal outlet temperature is T2+0.45 x ΔT=5ºC+0.45 x 60ºC = 32ºC.The resulting cooling water flow 4.4 kg/s. The optimal sizing of new coolers has been calculated for different temperatures on the process heat source that needs to be cooled. The total cost (including both operational and capital cost) per MWh cooling duty has been plotted against the cooling water exit temperature. The average inlet cooling water temperature is set to 15°C. The heat exchanger cost and heat transfer rate use in the calculations is based on a plate heat exchanger with water on the process side. 93 Figure 44: Total cooling cost vs. cooling water exit temperature It can be observed that the total operation cost increases asymptotically when the outlet temperature of the cooling water approaches the inlet temperature of 15°C. This is foreseen since the cooling water flow and related power consumption increases towards infinity. On the other end of the curve, the total operation cost increases as the cooling water temperature outlet temperature approaches the heat source temperature. The heat exchanger size and capital cost becomes too high. 94 VÄRMEFORSK K Drivers for reduced process water consumption and improved secondary heat systems Before analysing the possibilities of improving the secondary heat systems and reducing the water consumption it is important to establish the driving force for the specific plant. The drivers for reducing the process water consumption vary between different geographical locations depending on the recipient and fresh water availability. In Sweden we generally have good availability of clean fresh water but the firm environmental permits result in rather high costs for treating the exiting effluent. The driving force for reducing the cooling water consumption and possibly install cooling towers is mainly related to the fresh water availability during the seasons of the year. The drivers for optimizing the secondary heat systems are closely related to the economic incentives for reducing the primary energy consumption. The drivers for improved secondary heat systems vary considerable from mill to mill depending on the specific circumstances. There are some mills that suffer from an excess of steam during the summer time and internationally there are even examples of mills that are limited in power export possibilities because of undeveloped power distribution systems. In these cases there is obviously very limited driving force to improve the secondary heat system since there is no economical saving related to reducing the primary heat consumption. In other mills the cost for energy can be high due to capacity limitations in the boilers resulting in continuous consumption of expensive oil as support fuel. The most crucial situation is obviously if the plant production is limited by the steam generation capacity. The primary heat value can range from 0 SEK/MWh to around 1000 SEK/MWh assuming fuel oil as the most expensive support fuel. The same broad range can be observed for the power value. The same varying driving forces can be observed for reducing the process water consumption. Mills with good fresh water quality all year round, generously sized effluent treatment plant and allowance with respect to permits have a low economic incentive to further reduce process water consumption. Other mills located close to communities with firm environmental permits, coupled with poor fresh water source and low capacity margins in water treatment and effluent treatment plants, have a very strong driving force to reduce the process water consumption further. Typical process water and primary energy cost for Swedish conditions have been estimated in this appendix. 95 K.1 Costs related to water treatment and distribution The costs related to water production can principally be defined for four main water qualities:     Mechanically treated cooling water: Usage of cooling water with “mechanical” quality where the marginal cost is manly related to the electric power consumption for pumping water from water intake, through raw water treatment plant, mill distribution pumps and effluent. Cooling water from towers: Usage of cooling water taken and returned from a cooling tower where the marginal cost is related to the power consumption Chemically treated cooling water: Usage of cooling water with “chemical” quality where the marginal cost is mainly related to the chemical cost and sludge disposal in addition to the above power cost. Demineralized water: Chemically treated water where the ion content has been reduced significantly with ion exchange technology (IX) or reverse osmosis (RO) and electro deionization (EDI). Water is used as make-up for boilers and in some cases for production of special quality pulp such as dissolving pulp or market pulp with a low ion content. The cooling water cost varies very much between mills depending on the raw water quality, raw water treatment equipment and mill layout and distance to raw water intake and effluent treatment plant or discharge point. The production cost for chemically treated water also depends on the specific pulp and paper quality produced in the mill. Different technologies for water treatment plants in Sweden are discussed in Appendix B.3. The order of magnitude cost per m3 water consumption has been estimated within this study as rule of thumb for discussing optimization of the water systems. For comparison, the typical water cost for household water is 20 SEK/m3 in Sweden (however including production, distribution and effluent treatment). The water production cost is split into two parts, one including the pure operational cost (electricity, chemicals, maintenance) and a second including the capital cost of a hypothetical new treatment plant with 10 years depreciation. 96 VÄRMEFORSK Table 12.23: Typical average production costs of different water qualities. Mechanic- Water from Chemically ally treated cooling treated Deminerawater towers water lized water Specific perating cost (per required cooling demand) Specific capital cost on 10 years (per required cooling demand) Total specific cost (per required cooling demand) SEK/m3 0,02 0,08 0,67 SEK/MWh 0,76 2,60 23,0 SEK/m3 0,05 0,14 0,71 SEK/MWh 1,68 4,93 24,5 SEK/m 0,07 0,22 1,38 SEK/MWh 2,43 7,53 47,5 3 2,07 2,24 4,31 The production cost for chemically treated water includes the production cost for mechanical water. Furthermore, the cost for demineralized water includes the cost for chemically treated water. As noted above, the production cost for the different water qualities varies significantly from mill to mill. It is also important to understand that the above costs are average costs, and do not necessarily reflect the marginal cost of one additional consumed m3 water. As an example, if the final m3 water consumption forces the plant to upgrade the water treatment plant, the marginal cost becomes significantly higher since this investment is only allocated to the final m3. The specific water costs are based on an electric power cost of 500 SEK/MWh. Chemical costs for NaOH, Aluminum peratin, polymer and ClO2 used for producing standard chemically treated water are taken from some of recent projects. The specific cost per cooled MWh is based on a ΔT of 25°C for incoming and exiting cooling water. Higher or lower ΔT affects the electric power cost for water distribution. This is specifically discussed in chapter 7.3.4. The hypothetical capital cost per m3 is an order of magnitude cost and it is expected that it can vary considerable from case to case. The depreciation time of 10 years is longer than for other measures discussed within the report. Generally a longer financial depreciation time should be applied for more significant strategic investments. 97 K.2 Costs related to effluent treatment and discharge The costs related to effluent treatment can principally be defined into three main categories:    Cooling water bypasses the effluent treatment: Normally, uncontaminated cooling water by-passes the effluent treatment. There is no treatment cost related to this flow. The pump head in the water distribution system has sufficient pressure to transport the water to the recipient. The distribution cost is allocated to the water treatment plant. Cooling water mixed with process effluent: If clean cooling water is not separated from the process effluent it increases the volumetric load on the effluent treatment plant. Increased volumetric load reduces the retention time and performance. Also electric power is consumed for pumping the water in the effluent treatment. Process effluent: Process effluent containing COD and other pollutants is treated in the effluent treatment plant where electric power is consumed for pumping and aeration. Also consumption of chemicals such as nutrients for the biology and chemicals for precipitation increases with higher process effluent flow. The treatment cost varies mainly depending on the incoming COD content and composition, electric power cost and type of effluent treatment. The sludge disposal is accounted but the costs also vary considerably. Table 12.24: Typical average production costs of different effluents Process Clean water water without COD including etc. COD By-passed cooling water Specific perating cost Specific capital cost on 10 years Total specific cost SEK/m3 - 0.1 0.1-0.8 SEK/m3 - - 1.0 3 - 0.1 1.2-1.8 SEK/m As commented for the cost for water treatment and distribution in Appendix K.1, the effluent treatment cost varies significantly from mill to mill. It is also important to understand that the above costs are average costs, and do not necessarily reflect the marginal cost of one additional treated m3 water. The electrical power cost related to aeration and pumping of water represents the largest part of the operation cost. The specific water costs are based on an electric power cost of 500 SEK/MWh. Typical power consumption for effluent treatment plants in Sweden is 1 kWh/kg reduced COD. At COD concentration less than 0.8 kg/m3 the specific consumption increases [27]. 98 VÄRMEFORSK 3 Figure 45: Specific power consumption per m treated effluent in Swedish mills. The range of chemical costs for treating the effluent is taken from some existing plants and projects where a new effluent treatment has been designed. K.3 Costs related to primary heat savings in secondary heat systems The costs for steam and power have been defined within this study as the basis for the economical evaluation of different energy and water conservation options. This includes cost for external biomass, fuel oil and electrical power. The marginal steam cost has been calculated on this basis. The marginal energy cost varies considerably from mill to mill. The cost data below is assumed to be representative of a typical Swedish pulp and paper mill. K.3.1 Biomass fuel costs The marginal fuel cost for steam generation in Swedish mills is assumed to be represented by purchase of external biomass costing about 190 SEK/MWh based on the LHV on wet fuel. Boiler efficiency of 87% based on wet biomass and LHV results in a HP steam cost of 220 SEK/MWh. K.3.2 Electric power consumption The long term electric power cost is assumed to be 500 SEK/MWh in this study. The same power value is applied for sold and purchased power. This cost is thereby used 99 when evaluating any marginal increase or decrease in the mill power consumption. Presently the electric power cost is lower than 500 SEK/MWh but the long term trend is increasing energy costs. K.3.3 Electric power generation This study assumes that the pulp and paper mill turbine has enough capacity for cogeneration or power generation from all produced HP steam. The value of higher or lower cogeneration of electric power in turbines is set by electric power value of 500 SEK/MWh subtracted by the HP steam heat cost of 220 SEK/MWh for external biomass, yielding a net income of 280 SEK/MWh. Heat losses related to the electrical and mechanical efficiency of the generator are generally small and are neglected. The generated electric power can be granted subsidies (green certificates) if the plant is approved by the Swedish Energy Agency. The plant can receive anywhere from 0-100% subsidization of the generated renewable power and the value of each unit is presently about 200 SEK/MWh. The majority of the pulp and paper mills in Sweden are after 2012 not grated any subsidies. In this study no subsidies are accounted regarding cogeneration of electric power. K.3.4 Steam costs Generally generation of power in a thermal process means that heat is entering the power generator at a high temperature. Part of the heat entering the generating unit (turbo generator) is transformed to power, some part could be lost and the rest of the heat leaves the generating unit as heat at lower temperature than the entering heat. The heat leaving the generating unit can often be utilized for useful purposes, e.g. for pulp and paper process heat. That is what is meant by cogeneration. Consequently, the available heat in the high pressure steam can be used both for power and for heat demanding processes. If the heat leaving the power generating unit should not be utilized, it is no longer cogeneration. An example is condensing power generation. The generated power has to carry all the costs for the heat generation as the portion of the incoming heat that is led to the condensing section is ultimately lost to atmosphere via the cooling tower. How much power this means depends on the power generation process efficiency. Normally the efficiency is expressed as power generation (in kWh) relative to the net consumed heat (from the heat) leaving the generating unit (in MWh). This ratio (kWh/MWh) is often called the “alpha” value. A high alpha value means that for a given cogeneration basis, the possible power cogeneration is higher. In a cogeneration system it is possible to measure and calculate the efficiencies and the average alpha value in normal operation. However, when studying changes in the operation of a cogeneration system it is important to distinguish between the average 100 VÄRMEFORSK alpha value and the marginal value by a change. The marginal value normally could be much higher than the average. When calculating the economic impact of changes in the operation it is important to consider the marginal changes and to use marginal alpha and efficiency values. Marginal specific power generation (α) is estimated for typical operational conditions in Swedish mills. The resulting steam cost after the turbine after cogeneration of power based on external biomass as marginal fuel, is presented in table below. Table 12.25: Typical steam costs in Swedish pulp and paper mills. Parameter Unit [bar(a)] HP MP LP 60 15 5 Alfa (α) [kWh/MWh] 0 140 250 Steam cost [SEK/MWh] 220 180 150 The above alfa represents a turbine with relatively moderate power generation efficiency. 101 L Key data for bleach plant closure Table 26: Documented key data for evaluation of 2000 ADt/d the market pulp reference mill. REF Case A Case A leach Case B Case B leach Purges changed Purged ESP dust Lime mud kg/ADt kg/ADt 13.2 25.7 13.5 22.7 14.0 22.7 13.9 25.0 14.7 25.0 Make ups changed NaOH to white liquor CaO to lime kiln Increased ClO2 usage kg/ADt kg/ADt kg/ADt 8.5 12.1 0 7.6 10.8 0.18 8.1 10.8 0.18 5.8 11.9 0.18 6.7 11.9 0.18 Increased O2 usage Increased H2O2 usage kg/ADt kg/ADt 0 0 0.20 0.20 0.20 0.20 0.40 0.40 0.40 0.40 Water Fresh water usage in bleach plant Total amount of effluent Total COD in effluent m3/ADt m3/ADt kg/ADt 5.0 13.9 17.8 0.0 9.0 16.8 0.0 9.0 16.8 0.0 9.0 16.2 0.0 9.0 16.2 Cl/K kidney ESP dust from Recovery boiler to ash leaching Percent of ash to ash leaching H2SO4 usage kg/ADt % kg/ADt 0 0 0 0 0 0 22.4 12.9 0 0 0 0 35.2 19.7 0 Energy Fuel consumption in lime kiln, bark Heat generation in recovery boiler Evaporation plant steam consumption Fiber line steam consumption Total mill power consumption GJ/ADt GJ/ADt GJ/ADt GJ/ADt kWh/t 1.250 17.820 3.5 0.3 724 1.261 17.829 3.5 0.3 724 1.259 17.829 3.5 0.3 724 1.262 17.834 3.5 0.3 724 1.259 17.834 3.5 0.3 724 102 VÄRMEFORSK Table 27: Costs for chemicals, water and energy Chemicals NaOH cost (100%) CaO make-up SEK/ton SEK/ton 3510 1400 Increased ClO2 usage Oxygen cost Peroxide cost H2SO4 (usage in Cl/K kidney,100%) SEK/ton SEK/ton SEK/ton SEK/ton 10400 600 4000 1260 Process water Fresh water cost Effluent treatment cost SEK/m3 SEK/m3 0.7 0.8 SEK/MWh SEK/MWh SEK/MWh SEK/MWh SEK/MWh 190 220 150 150 500 Energy Fuel consumption in lime kiln Heat generation in recovery boiler Evaporation plant steam consumption Fiber line steam consumption Total mill power consumption 103