Transcript
Toxic Gas CiTiceLs®
Operating Instructions
Contents
Page
Introduction....................................................... 5TOX.2 Reaction Mechanisms ...................................... 5TOX.3 Selectivity......................................................... 5TOX.4 Continuous Operation ...................................... 5TOX.4 Emissions Monitoring ....................................... 5TOX.5 Start Up ........................................................... 5TOX.8 Operation ......................................................... 5TOX.8 Circuitry ........................................................... 5TOX.8 Calibration ..................................................... 5TOX.12 Cross Sensitivity Data .................................... 5TOX.13 In-Board Filters .............................................. 5TOX.13 In-Line SOx/NOx Filters .................................. 5TOX.14 Humidity Effects ............................................. 5TOX.15 Temperature Dependence .............................. 5TOX.16 Pressure Effects ............................................ 5TOX.17
Toxic Gas CiTiceLs Introduction The Toxic Gas CiTiceL development programme began in 1981 with the introduction of the Carbon Monoxide CiTiceL. Since then new CiTiceLs have been developed for various toxic gases, most recently ozone and ethylene oxide, resulting in a range of sensors with an enviable reputation for reliability, stability and robust design. These sensors are micro fuel cells, designed to be maintenance-free and stable for long periods. They use the technology that was pioneered with the original Oxygen CiTiceL, which results in a direct response to volume concentration rather than partial pressure. The central feature of the design is the gaseous diffusion barrier, which limits the flow of gas to the Sensing electrode. The electrode is therefore able to react all target gas as it reaches its surface, and still has electrochemical activity in reserve. This high activity reserve ensures each CiTiceL has a long life and excellent temperature stability.
Two electrode CiTiceLs The simplest form of sensor operating on electrochemical principles has two electrodes - Sensing and Counter - separated by a thin layer of electrolyte and connected by a low resistance external circuit. Gas diffusing into the sensor is reacted at the surface of the Sensing electrode, by oxidation or reduction, causing a current to flow between the electrodes through the external circuit. The current is proportional to the concentration of gas and can be measured across a load resistor in the external circuit. For reaction to take place the Sensing electrode potential must be within a specific range. As the gas concentration increases so does the current flow, causing a change in the potential of the Counter electrode (polarisation). With the electrodes connected together by a simple load resistor, the Sensing electrode potential follows that of the Counter. If the gas concentration continues to rise, the Sensing electrode potential will eventually move outside its permitted range. At this point the sensor will become non-linear, effectively limiting the upper concentration of gas a two electrode sensor can be used to measure.
Three electrode CiTiceLs The limitation imposed by Counter electrode polarisation can be avoided by introducing a third, Reference electrode, and using an external potentiostatic operating circuit. With this arrangement the Sensing electrode is held at a fixed potential relative to the Reference electrode. No current is drawn from the Reference electrode, so both maintain a constant potential. The Counter electrode is still free to polarise, but this has no effect on the Sensing electrode and so does not limit the sensor in any way. Consequently the range of concentrations a three electrode sensor can be used to measure is much greater. Most Toxic Gas CiTiceLs have a three electrode design. By controlling the potential of the Sensing electrode, the potentiostatic circuit also allows greater selectivity and improved response to the target gas. The same circuit is used 5TOX.2
Doc. Ref.: 5toxops.pmd Issue 4.1
July 2005
Figure TOX1 Three-electrode Toxic Gas CiTiceL schematic drawing
Capillary Diffusion Barrier O-Ring Seal Sensing Electrode Reference Electrode Counter Electrode
Electrolyte Reservoir
to measure the current flow between the Sensing and Counter electrodes. It can be a very small, low power device, and recommended circuits are given later in this chapter.
Four electrode CiTiceLs Further development of the three electrode design has led to four electrode sensors. These sensors have an additional working electrode, known as the Auxiliary. The signal from this electrode can be used to increase the selectivity of certain sensors, for example it is employed in the A5F sensor to null out hydrogen cross-sensitivity.
Reaction Mechanisms Gas diffusing into a CiTiceL is reacted at the sensing electode by oxidation (most gases) or reduction (e.g. nitrogen dioxide). Each reaction can be represented in standard chemical equation form. The oxidation of carbon monoxide, for example, at the sensing electrode can be represented by the equation: CO + H2O → CO2 + 2H+ + 2eThe counter electrode acts to balance out the reaction at the sensing electrode by reducing oxygen in air to water: ½O2 + 2H+ + 2e- → H2O A similar equation can be given for other CiTiceLs depending on the reaction of the gas they are designed for on the sensing electrode:
July 2005
Sulphur Dioxide (SO2) CiTiceLs:
SO2 + 2H2O → H2SO4 + 2H+ + 2e-
Nitric Oxide (NO) CiTiceLs:
NO + 2H2O → HNO3 + 3H+ + 3e-
Nitrogen Dioxide (NO2) CiTiceLs:
NO2 + 2H+ + 2e- → NO + H2O
Doc. Ref.: 5toxops.pmd Issue 4.1
5TOX.3
Selectivity CiTiceLs have been designed to be highly specific to the gas they are intended to measure, and the effects from other cross-interfering gases have been minimised. This is largely achieved by a combination of the following techniques:
1)
Development of specific electrode catalysts Reactions are catalysed by electrode materials specially developed by City Technology. Different gases have varying levels of electrochemical activity, as do electrode materials. In general, where a gas meets an electrode of a similar activity, it will be readily reacted. Although in practice it is impossible to prevent particularly active gases from reacting on least active electrode materials, in most cases gases outside the activity range of an electrode will not be reacted.
2)
Control of operating potential of the sensing electrode A major advantage of the three-electrode design is it allows a 'bias' voltage to be applied to the sensor, so enabling the oxidation or reduction of less electrochemically reactive gases. The appliance of such a bias promotes reactions which would not normally occur at the reference electrode potential.
3)
Use of chemical filters to remove interfering gases selectively A number of sensors have inboard filters to remove gases which would otherwise react on the sensing electrode. These filters vary in composition and size depending on the intended application and the likely exposure a sensor will have. For sensors designed for emissions applications, any inboard filters will have a high capacity. More details are given later in this chapter.
Continuous Operation All the CiTiceLs dealt with in this Section are designed, tested, and intended for use in discrete sampling applications. They are not designed for continuous emission monitoring. In applications where a continuous measurement is required CiTiceLs can only be used to measure up to 200 ppm CO, 100 ppm SO2, and 100 ppm NO. Higher concentrations may be measured intermittently but the actual concentration and the exposure time can only be determined by experiment. For this type of application further advice is available from City Technology's Technical Marketing Dept. The inboard filters in some of these sensors do not have the capacity to absorb gases continuously. These filters will be exhausted fairly rapidly by continuous use, or by exposure to concentrations of cross-interfering gases above 1000 ppm. In applications requiring removal of cross interfering gases, extra filter
5TOX.4
Doc. Ref.: 5toxops.pmd Issue 4.1
July 2005
capacity will be needed. i.e. through the use of outboard filters.
Emissions Monitoring The measurement of the gases emitted from the burning of fossil fuels in a boiler or similar appliance demands an awareness of complex technical issues when designing an instrument. There are two main areas for consideration:1) Sampling and cleaning of the gas; and 2) Use of the appropriate CiTiceL manufactured by City Technology.
1)
Sampling and Cleaning of the Gas The gas stream must be cooled to ambient temperature and cleaned through a suitable filter to remove solid material. The filter will need to be changed, so must be accessible, and must not remove acid gases (NOX/SO2). To avoid blockages and flooding, all condensed water must also be removed before it reaches the components of an instrument. (As a general rule the sample stream must be non-condensing and below 45°C before passing to the CiTiceL). It is also important to avoid pump pulsations produced in the gas stream reaching the CiTiceL, as this will cause enhanced signals. One of the major considerations in designing a sampling system is ensuring the active gas component of a gas stream does not adsorb on to the surfaces of the materials used in the system. This will deplete the concentration of gas, until the material is saturated. Surface adsorption is not usually a problem with CO and NO, but may affect SO2 and NO2. Adsorption effects may be minimized by employing materials with low absorption properties in contact with the gas, together with high flow rates and short gas lines. Fluoropolymers such as Polytetrafluoroethylene (PTFE), Trifluorinatedethylene (TFE) and Fluorinated Ethylene Propylene (FEP) have very low gas absorption properties and are suitable for use in gas handling systems. Stainless steel 316 and silicone rubber can be used as an alternative to plastic materials. Polyester is favoured for gases other than NO2. Once adsorbed, gases will desorb back into the gas stream when the system is purged with a clean gas. This will then be detected by the sensor until all
Figure TOX2 Suggested Gas Sampling Arrangement
Sample Probe
Water Trap
Particle Filter
O2
NO
Sensors
July 2005
Doc. Ref.: 5toxops.pmd Issue 4.1
SO2
Pump
In-line SOx\NOx Filter
NO2
CO
5TOX.5
surface adsorbed gas is removed. All CiTiceLs must be suitably protected by appropriate filtration to remove highly corrosive constituents and particulate matter. In most cases no external protection over the gas entry holes is provided with the sensor. Protection is therefore required to prevent undesirable ingress by solids and liquids (such as would be encountered in a flue gas stream). At the same time the gas entry holes must be kept free of obstruction to ensure the correct calibration of the CiTiceL. Condensing conditions at the CiTiceL must always be avoided.
2)
Choice of Appropriate CiTiceL CiTiceLs enable the accurate measurement of a range of gases encountered in emissions applications. They have a two or three year minimum operating life under normal operating conditions. Each sensor is packaged in a purpose made, corrosion resistant, plastic moulding and the 5-series sensors feature replaceable filters for CO and NO. The gases most commonly of interest are carbon monoxide, sulphur dioxide, nitric oxide and nitrogen dioxide and are dealt with below. For applications involving the measurement of other gases, please contact City Technology.
5-series sensors The 5-series range of CiTiceLs feature an easy bayonet mounting system. All sensors in the range, including oxygen, are available in this compact, functional hardware. CO and NO 5-series sensors feature replaceable filters which make them more suitable for use in applications which contain high levels of crossinterfering gases. The replaceable filter can be easily and regularly changed in applications where inboard filters would normally become exhausted. Replaceable filters therefore extend the useful operating life of the sensor. Carbon Monoxide type :
5F CiTiceL (Nominal Range: 0-4000ppm)
Standard Carbon Monoxide CiTiceLs have a cross sensitivity to SO2 and NOx. To remove high concentrations of these gases before they reach the sensing electrode the 5F CiTiceL has an inbuilt chemical filter. The capacity of this filter is limited, but may, however, be replaced. Filter replacement is a unique feature of the 5-series CO sensors. Carbon Monoxide type:
A5F CiTiceL (Nominal Range: 0-2000ppm)
The A5F is designed to meet TÜV specifications for gas sampling instruments. It is ideal for measuring carbon monoxide in emissions applications where hydrogen may be present. Performance is enhanced by an additional auxiliary electrode for hydrogen compensation and a high capacity, replaceable, inboard filter for NOx and SO2 removal.
5TOX.6
Doc. Ref.: 5toxops.pmd Issue 4.1
July 2005
Carbon Monoxide type :
5MF CiTiceL (Nominal Range: 0-40000ppm)
Standard Carbon Monoxide CiTiceLs have a cross sensitivity to SO2 and NOx. To remove high concentrations of these gases before they reach the sensing electrode the 5MF CiTiceL has an inbuilt chemical filter. The capacity of this filter is limited, but may, however, be replaced. Filter replacement is a unique feature of the 5-series CO sensors. Sulphur Dioxide type:
5SF CiTiceL (Nominal Range: 0-2000ppm)
Chemical filters are unnecessary as this sensor has low cross-sensitivity to other common flue gases. Nitric Oxide type:
5NF CiTiceL (Nominal Range: 0-2000ppm)
The 5NF features a high capacity replaceable filter for SO2 and NO2 removal. This sensor has been developed to meet TÜV requirements for NO measuring instruments and also includes a fourth, scavenging electrode for enhanced performance. Nitrogen Dioxide type:
5ND CiTiceL (Nominal Range: 0-100ppm)
A high range NO2 sensor with low cross-sensitivity to CO, SO2 and NO which completes the 5-series range.
July 2005
Doc. Ref.: 5toxops.pmd Issue 4.1
5TOX.7
Toxic Gas CiTiceL Operation Start Up To maintain the CiTiceL in a 'ready to work' condition, most sensors are despatched from City Technology with a shorting link connected across the sensing and reference terminals. This link must remain in place during storage and should only be removed when the CiTiceL is used in an instrument. The electrodes must be reshorted when the instrument is switched off, otherwise long start up times will result. In the recommended circuit opposite this is done using a shorting J-FET, which keeps the electrodes shorted whenever the circuit is unpowered.
Operation For correct operation, CiTiceLs require a small supply of oxygen to the counter and reference electrodes. This is usually provided in the sample stream, by air diffusing to the front of the sensor, or by diffusion through the sides and rear of the sensor (a few thousand ppm is normally sufficient). Continuous exposure to an anaerobic sample gas may cause the sensor to malfunction in spite of the rear oxygen access paths. The sensor must not therefore be completely potted with resin or totally immersed in an anaerobic gas mixture.
Circuitry a) Standard three electrode Toxic Gas CiTiceLs Figure TOX3 shows the recommended circuit for use with standard three electrode CiTiceLs. The output from the circuit will be Positive with respect to common for CO and SO2, which are oxidised, and Negative with respect to common for NO2 which is reduced at the sensing electrode. The circuit is designed for 'unbiased' operation as recommended for CO, SO2, and NO2. CiTiceLs for NO require biased operation for which additional circuitry is needed. This is described in more detail in the next section. The function of the counter electrode is to complete the electrochemical circuit and its potential relative to the sensing and reference electrodes is not fixed by the circuit. Under quiescent conditions, the cell is drawing a very small current and the counter electrode will only be a few mV more negative than the reference. When gas is detected, the cell current rises and the counter electrode polarises negative with respect to the reference electrode (positive for NO2). While the cell current stabilises very quickly, the counter electrode polarises quite slowly and may continue to drift even though the sensor signal is stable. This is quite normal, and in practice the maximum counter electrode polarisation likely to be met is 300-400mV with respect to the reference 5TOX.8
Doc. Ref.: 5toxops.pmd Issue 4.1
July 2005
electrode. In practical terms this means the circuit ground should be derived at a higher value than the negative supply rail (e.g. 1V), so that IC1 can give a negative output. The voltage developed across RL should be restricted to less than 10mV under all conditions otherwise sensor performance will suffer. Keeping RL low also ensures a faster response time, and although in this circuit it can be reduced to zero, a small finite value gives a better balance between circuit noise and response time. Figure TOX3 Toxic Gas CiTiceL: Unbiased Inverting Amplifier Circuit IC1 - This amplifier should have either a low offset or have its offset nulled out. The PMI OP77 and OP-90, Intersil or Teledy ne 7650, and Linear Technolog y LT1078 are all suitable. IC2 - This amplifier acts as a current to voltage converter and its offset per formance is less critical. The OP-77 or similar is a suit able choice Recommended values of Rload are given on data sheets.
b)
For a CiTiceL to remain in a ‘ready to work’ state when an instrument is switched off, the reference and sensing electrodes must be shorted together. This is done by shorting the reference to the circuit common with either a ganged on/off switch or an FET as shown above. While shorted it is important to avoid exposure to active gases or solvent vapours.
Nitric Oxide CiTiceLs - "Biased Operation" Nitric Oxide CiTiceLs are designed to work with the sensing electrode potential above that of the reference electrode, otherwise known as 'biased' operation. Circuitry required is basically the same as shown in figure TOX3, but is modified to provide a bias voltage by placing the positive input of IC1 at the required potential below the circuit common. The modified circuit is shown opposite. The bias voltage must be applied via IC1 so as not to draw any current from the reference electrode. It must not be applied by directly connecting a battery to the reference and sensing electrodes. It is strongly recommended that the bias voltage be maintained at all times, even when an instrument is switched off. If the bias potential is not maintained, very long start up times will result when the instrument is switched on. The recommended bias voltage for Nitric Oxide CiTiceLs has the sensing electrode 300mV more positive than the reference electrode. This is different
July 2005
Doc. Ref.: 5toxops.pmd Issue 4.1
5TOX.9
Figure TOX4 Standard Circuit for Biased Operation CiTiceLs IC1 - This amplifier should have either a low offset or have its offset nulled out. The PMI OP77 and OP-90, Intersil or Teledyne 7650, and Linear Technolog y LT1078 are all suitable. IC2 - This amplifier acts as a current to voltage converter and its offset per formance is less critical. The OP-77 or similar is a suit able choice V ref is a precision bandgap voltage reference. The Teledyne TC04 series is recommended. Recommended values of Rload are given on data sheets. Figure TOX5 Recommended Circuit for 5NF CiTiceL
Counter
0µ1
10K
Reference
V-
10K -
IC2 - This amplifier acts as a current to voltage converter and its offset per formance is less critical. The OP-77 or similar is a suitable choice V ref is a precision bandgap voltage reference. The Teledyne TC04 series is recommended.
5TOX.10
10K
RGain
Sensing
22K + IC2
RLoad
Vout
Set Bias
10K
Vref 10K
10K
from previously recommended figures, but has been shown to offer the greatest balance of features for operational use. With this new level of bias, sensors no longer suffer from the problem of very large baselines when operated at temperatures towards the top end of the operating range. Note: The reference and sensing electrodes of CiTiceLs requiring biased operation are not meant to have the same potential, so are despatched from City Technology without the usual shorting link. As shorting can cause permanent damage, these CiTiceLs must be stored with the electrodes unshorted. For this reason the shorting FET used in the unbiased circuit is omitted in the recommended bias circuit. Please note that some sensors can be shipped with a bias board for transit rendering them available for immediate use. Please refer to specification section.
V-
0µ1
Counter Reference
10K
10K
-
Scavenging
IC1 - This amplifier should have either a low offset or have its offset nulled out. The PMI OP77 and OP-90, Intersil or Teledyne 7650, and Linear Technolog y LT1078 are all suitable.
IC1
+
IC1
10K
+ RGain
Sensing
22K
RLoad
+
IC2
Set Bias
10K
Vout Vref
10K
10K
The 5NF CiTiceL incorporates a fourth, "scavenging," electrode which should be connected to the instrument ground. In a normal 3-electrode sensor the products of NO oxidation can interfere with the potential of the sensor's reference electrode and possibly cause non-linearity of the signal and non-repeatability. This is especially the case in applications where the sensor is exposed repeatedly to high concentrations of nitric oxide. The "scavenging" electrode helps overcome this problem by chemically reacting the products of nitric oxide oxidation.
Doc. Ref.: 5toxops.pmd Issue 4.1
July 2005
c) A5F CiTiceLs - "Hydrogen Compensated" The performance of three electrode Carbon Monoxide CiTiceLs can be enhanced by the introduction of a second working electrode, known as the auxiliary. Mounted behind the sensing electrode, the fourth electrode gives a signal predominantly in response to the hydrogen content of a test gas. This has been done with the A5F CiTiceLs. An example of a simple circuit which may be used to run these sensors is shown below. This is designed to hold both the sensing and auxiliary electrodes at 250mV above the reference electrode, a mode of operation known as ‘biasing’ (see above). Operating in this mode reduces the magnitude of the coefficients which determine the hydrogen response (b and d in the calculations shown overleaf), and the properties of the sensor are much improved. It means however the circuitry which powers the sensor and provides the bias must be continuously switched on to avoid long start-up times. Figure TOX6. A5F Circuit Diagram IC1 - This amplifier should have either a low offset or have its offset nulled out. The PMI OP77 and OP-90, Intersil or Teledy ne 7650, and Linear Technolog y LT1078 are all suitable. IC2, IC3 These amplifiers act as current to voltage converters and their offset performance is less critical. The OP77 or similar is a suitable choice
1K
0µ1
Counter Reference
1K8
10K
10K
-
Auxiliary
IC1 +
15K
10K
Sensing 10R
AUXILIARY
+
68K
IC2
10K Vref = -1.26V 10R
SENSING
+
IC3
V ref is a precision bandgap voltage reference. The Teledyne TC04 series is recommended.
When operating without bias the sensor may be used immediately after switch on providing that the sensing and auxiliary terminals have been shorted to the reference terminal while the instrument was ‘off’. A J-FET, such as the one used in the circuit shown in figure TOX3, is suitable.
Recommended values of Rload are given on data sheets.
A typical flue gas can contain both carbon monoxide and hydrogen. When an A5-Series sensor is exposed to this mixture the sensing electrode produces a signal proportional to both gases, and the auxiliary produces an output which is mostly due to the hydrogen present. This design, together with an inboard chemical filter for the removal of NO, NO2, H2S and SO2 allows much more accurate measurement of the carbon monoxide content of flue gas than otherwise possible. The equations required to calculate the carbon monoxide concentration of a sample from the two signals are given on the next page.
July 2005
Doc. Ref.: 5toxops.pmd Issue 4.1
5TOX.11
The output from an A5F CiTiceL can be expressed as follows:Let:
IS IA [CO] [H2]
= = = =
Sensing electrode signal; Auxiliary electrode signal; Concentration of CO in the mixture; Concentration of H2 in the mixture.
Then:
IS IA
= =
a[CO] + b[H2] .............................. (1) c[CO] + d[H2] .............................. (2)
Eliminating [H2] from (1) and (2) gives:[CO] =
(ISd-IAb) / (ad-bc) ........................ (3)
If a separate calibration is done for carbon monoxide and for hydrogen the coefficients a,b,c and d may be measured using equations (1) and (2). Equation (3) may then be used to derive the actual carbon monoxide concentration in a mixture of carbon monoxide and hydrogen. As part of the sensor design parameters it is important to make ad-bc as large as possible (see equation 3). The current sensor design has the value of c between 1% and 10% of a, and the ratio b/d lies between 1/2 and 5/2. Generally coefficients a, b, c, and d will also be functions of temperature, so a more accurate expression of equations (1) and (2) will be: IS IA
= =
a(T)[CO] + b(T)[H2] .................... (4) c(T)[CO] + d(T)[H2] .................... (5)
To solve this system the carbon monoxide and hydrogen calibrations must be done at more than one temperature and curve fitting techniques or look-up tables employed. Generally a microprocessor-based solution is to be recommended. Further information on the temperature behaviour of the A5F CiTiceL is available from City Technology.
Calibration The patented capillary barrier design makes CiTiceLs extremely stable, so loss of calibration is a gradual process, and routine recalibration is only necessary about every six months. For maximum accuracy, CiTiceLs should be calibrated using a gas mixture in the range where most measurements are to be made. Where this is not possible, a mixture towards the top of the CiTiceL range should be chosen. Calibration gases exceeding the range of the CiTiceL must not be used as this may not provide an accurate calibration. For most emissions monitoring applications aerobic mixtures approaching 1000ppm for CO and SO2 offer the greatest accuracy. However safety considerations may require the use of lower concentrations of these gases and 50-200ppm CO and SO2 is acceptable. NO must always be in an anaerobic mixture and 500ppm is more usual. The table below gives gas concentrations and flow rates appropriate for calibrating each CiTiceL, providing optimum performance with the minimum of gas hazard.
5TOX.12
Doc. Ref.: 5toxops.pmd Issue 4.1
July 2005
Since calibration normally only involves exposing the sensing face of the CiTiceL to gas for a relatively short time, a calibration gas need not contain oxygen - sufficient is supplied from ambient air through the side and back access paths for a limited period. For most purposes a five minute exposure time is satisfactory to achieve a stable calibration signal. Depending on the equipment used, NO2 CiTiceLs may need a longer exposure time due to the effects of surface adsorption. Gas
CiTiceL
Gas Concentration
Minimum Flow Rate
5F, A5F, 5MF
1000ppm
150mls/min
Sulphur Dioxide
5SF
100ppm
150mls/min
Nitric Oxide
5NF
100ppm
150mls/min
Nitrogen Dioxide
5ND
100ppm
150mls/min
Carbon Monoxide
Cross Sensitivity Data Data Sheets includes tables of the cross-sensitivity of each CiTiceL to gases other than their target gas (a full table covering all 5-Series sensors is given in Appendix 6). Depending on the nature of the reaction each gas has with the sensor, the effect can either decrease the signal (negative cross-sensitivity) or increase the signal (positive cross-sensitivity). Each figure represents the reaction of the sensor to 100ppm of gas, so providing a percentage sensitivity to that gas relative to its target gas. If a sensor shows cross-sensitivity to a particular gas, whether or not this is a threat to accuracy in an application depends on the degree of accuracy required. For instance where ±10% accuracy is needed, any gas likely to be present in a high enough concentration to cause a 10% 'error' signal should be monitored separately. CiTiceLs can be assumed to respond linearly to cross-interfering gases. Although the tables provide a guide they do not dictate the behaviour of any particular sensor, and any sensor could easily exhibit a cross-sensitivity 1015% above or below the stated figure. Sensors may also behave differently with changes in ambient conditions.
In-Board Filters Emissions monitoring CiTiceLs for carbon monoxide and nitric oxide have been designed to eliminate cross sensitivities to other gases which may be present by the use of chemical, inboard filters. These filters are designed to remove certain gases from the gas stream before it reaches the sensing electrode, so eliminating particular cross-sensitivities. July 2005
Doc. Ref.: 5toxops.pmd Issue 4.1
5TOX.13
The life of the filter material is limited, although sufficient filter material is used to last the sensor life in normal sampling operations. However in order to meet the requirements of the most demanding applications, the 5-series range has been developed. The 5-series CO and NO sensors feature high capacity filters which are easily replaced.
In-Line SOx/NOx Filters If a particular application involves unusually high levels of cross-sensitive gases, it may be desirable to supply additional filter capacity to protect a CiTiceL's internal filter. This is particularly relevant in emissions applications involving heavy oils or woodburners which liberate high levels of nitrogen dioxide, nitric oxide and sulphur dioxide. These gases can cause problems when measuring carbon monoxide. City Technology has developed an in-line filter for use with carbon monoxide CiTiceLs in pumped systems. The filter is designed to remove the SO2, NO and NO2 in a sample gas without affecting the carbon monoxide concentration. These cartridge filters are inserted in the gas sampling system to absorb acid gases which may give rise to a false signal. The filter is intended for use only with the CO sensor, to absorb SO2, NO, and NO2, and is easy to replace when exhausted, (as indicated by a colour change).
(a) Construction The dimensions of the filter are shown below. The housing is constructed from a modified polyamide material, selected for its high strength and resistance to chemical attack. However, it is not designed to be used in high concentrations of the following substances: Alcohols Amines
Ketones Pyridine
Phenols Chlorinated Solvents
The spigots of the filter can accommodate 6mm flexible tubing for connection purposes, provided the pressure of the gas line does not exceed three atmospheres.
Figure TOX7 In-line SOx/NOx Filter: Outline drawing
4.5 mm
87.0 mm
26.5 mm
1.5 mm
47.0 mm 6.5 mm
All Tolerances ±0.15mm
5TOX.14
Doc. Ref.: 5toxops.pmd Issue 4.1
57.0 mm
July 2005
(b) Filter Life Figure TOX8 shows the removal capacity of the filter with a 1 litre/min gas stream containing NO, NO2, and SO2, each at 1000 ppm. In practice the filter life will vary according to the application and will depend on many other operational factors. When new, the filter material is purple, and as it gradually loses its capacity to remove acid gases, the colour of the material slowly changes to a dark brown. Figure TOX8 In-line SOx/NOx Filter: % Efficiency (1000ppm at 1litre/min)
Filters are supplied in packs of ten with protective caps to prevent contamination of filter material during transit and storage.
Order Code In-line SOX/NOX Filter .............................................. B100
Humidity Effects Toxic Gas CiTiceLs are based on the use of aqueous electrolytes which, in conjunction with the porous diffusion barrier, permits water vapour to be absorbed into the electrolyte under conditions of high water vapour pressure, and allows the electrolyte to dry out at very low ambient water vapour pressure. Provided conditions are non-condensing the performance is relatively unaffected by humidity and will simply follow the change in concentration of the measured gas which results from changes in humidity. However when rapid changes in humidity occur, the sensors will show a transient response which should die away after about 20-30 seconds. Continuous operation is possible between 15% and 90% RH over the full temperature range, as the electrolyte will reach an equilibrium with the external water vapour pressure, at a volume and concentration which does not affect the sensor's life or performance. Outside these conditions, water transfer may occur. Under continuous operation at high temperatures and 90-100% RH, water will slowly diffuse in. However water uptake is only harmful when the liquid volume increase exceeds the free space available. When this happens the sensor becomes prone to leakage - increasingly so as more and more water is taken up by the sensor. July 2005
Doc. Ref.: 5toxops.pmd Issue 4.1
5TOX.15
Removal from this humidity to a lower RH before leakage occurs will gradually restore the sensor to its original condition and no permanent harm will result from this exposure. Similarly in continuous operation at 0-15% RH water will diffuse out. This will only be a problem when the volume of electrolyte has decreased by more than 40%, at which point the sensor gas sensitivity will be affected and the housing and seals may be attacked by the very concentrated electrolyte. Provided a sensor is not left in this condition long enough for such a reduction in the electrolyte to take place, it can be restored by exposure to a RH above 15%. At this level water ingress will begin to restore the water balance. The rate of transfer depends on the ambient temperature and relative humidity at the sensor. It also depends on the electrolyte and capillary hole size, both of which vary from one type of CiTiceL to another. A medium sensitivity sensor will continue to operate for around six to seven weeks at 100% RH and 40°C (continuous) and two to three weeks at 0% RH (continuous). Emissions monitoring CiTiceLs are relatively low sensitivity sensors, and will in general have slower water transfer rates and can be used for longer periods.
Temperature Dependence Both the span signal and the baseline (zero gas current) are affected by temperature.
(a) Span The output from a CiTiceL will vary only slightly with temperature. The temperature coefficient graphs in Part I specifications (example opposite) show how the output of each sensor will change with gradual shifts in temperature. These graphs show the typical variation in span output with temperature for CiTiceLs calibrated at 20°C to a reading of 100% from a suitable test gas. Each graph is based on a sample of about 16 sensors (unless stated otherwise) providing a high statistical probability that the behaviour of most CiTiceLs of that type will fall within the range of +3 to -3 times the standard deviation. Figure TOX9 Typical Temperature Data
Example: 5F
5F CiTiceL Temperature Coefficient Data
110 100
Sample
90 80 70 Mean
+3sd
-3sd
60 50 -40
5TOX.16
-20
0 20 Temperature (°C)
Doc. Ref.: 5toxops.pmd Issue 4.1
40
60
July 2005
Note: For rapid changes in temperature a transient response will be seen, which dies away after about 20-30 seconds, the output then settles at the new level according to the graph data.
(b) Baseline The baseline signal follows an exponential relationship with temperature change. The baseline approximately doubles for every 10°C increase in temperature.
Pressure Effects CiTiceLs give a transient response when exposed to a sudden change in pressure in the presence of a measured gas. The peak signal decays in only a few seconds. This can be a particular problem when using sampling pumps, as they may introduce pressure fluctuations into the gas stream. Pressure pulsations can be avoided by ensuring the sensor is positioned at the atmospheric end of the sample train. Alternatively a flow restrictor placed upstream from the CiTiceL will also help to damp out pressure oscillations. Another effective measure is to ensure that the back pressure downstream from the CiTiceL is effectively zero, so allowing an unrestricted flow of gas to ambient air. However it is important to prevent back diffusion from the ambient air diluting the gas stream and lowering the gas concentration being measured. Back diffusion can be reduced for example by the provision of an exhaust gas tube 4mm internal diameter and 8mm length.
July 2005
Doc. Ref.: 5toxops.pmd Issue 4.1
5TOX.17
5TOX.18
Doc. Ref.: 5toxops.pmd Issue 4.1
July 2005
