Carbon dioxide and/or hydrogen sulphide detection system and method and use thereof

Abstract

Various embodiments of the present disclosure are directed to carbon dioxide and/or hydrogen sulphide sampling and detection system and method for determination of the content of gaseous CO2 and/or H2S in a liquid, among other chemical compounds. In one embodiment, the detection system includes a membrane block having a liquid sample inlet port and a sample outlet port between which a sample flow path extends. The membrane block includes a first membrane unit and a second membrane unit. The first membrane unit includes a sample flow on the first side of a first permeable membrane element, and a carrier gas flow on the second side of the first permeable membrane element. The second membrane unit having a sample flow on the first side of a second permeable membrane element and a carrier gas flow on the second side of the second permeable membrane element.

Claims

1. A carbon dioxide or hydrogen sulphide sampling and detection system for determination of the content of gaseous carbon dioxide or hydrogen sulphide in an aqueous liquid, or for the detection of the total carbonate content in an aqueous liquid, wherein the sampling and detection system comprises: a membrane block including a liquid sample inlet port, a sample outlet port, a liquid sample flow path that extends between the liquid sample inlet port and the sample outlet port, a first membrane unit having a sample flow on a first side of a first permeable membrane element, and a carrier gas flow on a second side of the first permeable membrane element, and a second membrane unit having a sample flow on a first side of a second permeable membrane element and a carrier gas flow on a second side of the second permeable membrane element, and a carbon dioxide gas sensor or a hydrogen sulphide detection unit; and a gas circulation means; wherein said first and second membrane units are arranged in series in the liquid sample flow path; and wherein a gas flow path is a closed loop that includes the gas circulation means, the second side of the first membrane unit, the second side of the second membrane unit, and the carbon dioxide gas sensor or the hydrogen sulphide detection unit, and wherein the closed loop gas flow path further includes a breather valve arrangement with two serially connected three way valves between the carbon dioxide gas sensor or the hydrogen sulphide detection unit and atmosphere.

2. The system according to claim 1, wherein the carbon dioxide sensor is a sensor based on IR technology.

3. The system of claim 2, wherein the carbon dioxide gas sensor is configured and arranged to detect an absorption in the infra-red spectrum.

4. The system according to claim 1, wherein the hydrogen sulphide detection unit includes an electrochemical measuring cell with an electrolyte, a measuring electrode, a counter electrode and a reference electrode.

5. The system of claim 1, wherein the first and second membrane elements are hydrophobic membranes, the hydrophobic membranes are selected from the group consisting of: poly tetrafluorethylene membranes, poly dimethyl siloxane membranes and combinations thereof.

6. The system of claim 1, wherein the membrane elements have a pore size less than 0.02 microns.

7. The system of claim 1, further including a set of bypass valves and a bypass loop with a mixer station configured and arranged for admixing one or more acids into the liquid flow path.

8. The system of claim 1, wherein the system is configured and arranged to be portable.

9. The system of claim 1, wherein the system is configured and arranged to be operated in an in-line configuration within a fish farm.

10. The system of claim 1, wherein the sampling and detection system is configured to determine the gaseous content of gaseous carbon dioxide or hydrogen sulphide from a liquid broth.

11. A method for sampling and detection of carbon dioxide or hydrogen sulphide in a liquid, the method comprising the steps of: isolating gaseous carbon dioxide or hydrogen sulphide from the liquid in a membrane block by a sample flow liquid passing through the membrane block, bypassing the gaseous carbon dioxide or hydrogen sulphide contained in the sample flow through first and second permeable membrane elements in first and second membrane units of the membrane block and into a gas flow, while maintaining a sample liquid flow in the sample flow, where said first and second membrane units have the sample flow on the first side of first and second permeable membrane elements and a carrier gas flow on the second side of the first and second permeable membrane element, and said first and second membrane units are arranged in series in the liquid sample flow, and wherein the gas flow is a closed loop, and the gas is circulated through the membrane units and to a carbon dioxide gas sensor or a hydrogen sulphide detection unit arranged in the gas flow, and determining at least one of the content of gaseous carbon dioxide or hydrogen sulphide in the liquid, or detecting the total carbonate content in the liquid, wherein the gas flow includes a breather valve arrangement with two serial connected three-way valves arranged between the carbon dioxide gas sensor or the hydrogen sulphide detection unit and the gaseous side of the membrane block gas flow path, and further between the carbon dioxide gas sensor or the hydrogen sulphide detection unit and an air pump, and wherein the method further includes the steps of opening the breather valve arrangement between subsequent measurements for providing a connection from the gas flow path to ambient atmosphere, and allowing humidity to leave the carbon dioxide sensor or the hydrogen sulphide detection unit or for venting the gas present in the gas flow to the atmosphere prior to a subsequent measurement of carbon dioxide or hydrogen sulphide.

12. The method of claim 11, wherein the gas flow extends through the first and second membrane units in counter-current direction relative to the liquid sample flow.

13. The method of claim 11, wherein the gas flow extends through the first and second membrane units in concurrent direction relative to the liquid sample flow.

14. The method of claim 11, wherein the liquid is an aqueous liquid and further including the step of adding one or more acids to the aqueous liquid sample for setting free carbon dioxide from the aqueous liquid sample, prior to measuring the free carbon dioxide, and thereby obtaining a measure for the total carbonate concentration in the aqueous liquid sample.

15. The method of claim 11, characterized in, raising the temperature of the liquid sample flow to 25-45° C. or maintaining the temperature in the membrane block at 25-35° C.

16. The method of claim 11, wherein the liquid is an aqueous liquid.

17. The method of claim 11, further including raising the temperature of the liquid sample flow to around 30° C., or maintaining the temperature in the membrane block at around 30° C.

18. The method of claim 11, wherein the gas flow includes a breather valve arrangement with two serial connected three-way valves arranged between the carbon dioxide gas sensor or the hydrogen sulphide detection unit and a gaseous side of the membrane block gas flow path, and also further between an air pump and an air inlet to the membrane block, the method further including the steps of opening the breather valve arrangement between subsequent measurements for providing a connection from the gas flow path to ambient atmosphere, and allowing humidity to leave the carbon dioxide sensor or the hydrogen sulphide detection unit or for venting the gas present in the gas flow to the atmosphere prior to a subsequent measurement of carbon dioxide or hydrogen sulphide.

Description

DESCRIPTION OF THE DRAWING

(1) The present invention will in the following be described in more detail with reference to the figures in which

(2) FIGS. 1a-1c show the CO2 or H2S detection system according to the present invention from one side, from above and from the opposite side,

(3) FIG. 2a is a cross sectional view of the membrane block of the system,

(4) FIG. 2b is a side view of the membrane block with the liquid flow ports of the first and second membrane units,

(5) FIG. 3 shows the breather valve arrangement in the gas loop,

(6) FIGS. 4a-4b show the sample preparation with acid addition for detection of dissolved carbonates and/or total carbonate content,

(7) FIGS. 5a-c show graphs from the tests performed in examples 1-3,

(8) FIG. 6 shows a combined CO2 and H2S detection system according to a variant of the present invention seen from above,

(9) FIG. 7a shows a variant where the two membrane elements of the membrane blocks are arranged side by side arrangement,

(10) FIG. 7b shows a variant where the two membrane blocks are arranged each with their own air loop and sensors are applied in series,

(11) FIG. 8 shows the arrangement of a H2S detection unit in a gas loop, where the H2S sensor is provided with a distribution cap, and

(12) FIG. 9 graph from the tests performed in example 4.

DETAILED DESCRIPTION OF THE INVENTION

(13) FIG. 1a-1c shows the main components of the carbon dioxide and/or hydrogen sulphide detection and monitoring system 1. The carbon dioxide and/or hydrogen sulphide detection and monitoring system 1 for liquid aqueous samples is designed initially to extract CO2 gas and/or H2S gas out of a liquid sample flow to a gas loop 10. This is achieved in the membrane block 2.

(14) Initially, the liquid flow sample is filtered in a (not shown) filter unit in order to remove any particulate matter. Any standard particulate filter screen would be applicable as long as it is able to remove particles down to a size of approx. 10 μm.

(15) In addition, initial removal of free bubbles may be carried out using a closed container (not shown in drawings) with an airspace in which free bubbles in the sample flow may be collected, and by extracting the liquid sample flow from the bottom of the container.

(16) The membrane block 2 comprises a liquid sample flow inlet 3 and a liquid sample flow discharge port 4. The liquid sample flow direction is illustrated by the arrows 5,6 in FIG. 1a.

(17) The membrane block 2 comprises two separate membrane units 2a, 2b. The membrane units 2a, 2b are arranged in series by providing a flow connection 7 between an outlet port 8 of the first membrane unit 2a with an inlet port 9 of the second membrane unit 2b.

(18) The gas flow loop 10 has gas flow outlet 11 from the gas side 21 of the membrane 19 in the first membrane unit 2a of the membrane block 2. The gas loop 10 further has an inlet 12 to the gas side 21 of the membrane 19 in the second membrane unit 2b.

(19) From the gas flow outlet 11 of the membrane block 2, the gas flow circulates through the gas loop 10.

(20) The breather valve assembly comprising a first 13 and a second three-way valve 14 are included in the gas flow between the gas flow outlet 11 and the CO2 sensor 15. Alternatively, if the system is used for H2S detection only, then a H2S detection unit 29 (not shown in FIG. 1) is provided instead of the CO2 sensor 15.

(21) An air pump 16 is arranged on the downstream side of the sensor 15 as shown in FIG. 1b. The air pump 16 may, however, in alternatives be installed upstream of the sensor 15 or between the gas flow outlet 11 from the membrane block and the valve 13.

(22) Alternatively, the breather valve assembly may be arranged between the air pump 16 and the sensor 15, or between the air pump 16 and the membrane block gas flow inlet 12.

(23) The liquid sample flow inside the membrane block 2 (see FIG. 2a) is spread evenly across a hydrophobic membrane 19 on the liquid flow path 18 in the first membrane unit 2a. The gas diffuses through the membranes to the gas loop 10, driven by the gas partial pressure gradient towards equilibrium as per Henry's Law. The liquid sample flow then passes through the tube connection to the liquid flow path 18 in the second membrane unit 2b again spreading evenly across the second hydrophobic membrane 19 in the second membrane unit 2b, again with gas diffusion across the membrane. CO2 gas and/or H2S gas can diffuse either way across the membranes.

(24) The pressure in the liquid sample flow is preferably raised about 100-300 mbar relative to atmospheric pressure, or more preferred 150-250 mbar relative to atmospheric pressure or more preferred 200 mbar relative to atmospheric pressure.

(25) Preferably, the pressure in the liquid sample flow is preferably minimized. Indeed the liquid can be sucked through the membrane block 2 using a vacuum pump arranged at the liquid sample outlet of the membrane block 2. This reduces the pressure in the liquid sample flow to slightly below atmospheric pressure, e.g. 50-25 mbar below atmospheric pressure, without causing significant reduction of the transfer of gas across the membranes in the membrane block.

(26) The air pressure in the air loop varies with the content of CO2 and/or H2S gas in the sample. Usually, the pressure in the air loop is only raised slightly above to atmospheric pressure during measurements, such as between 20-60 mbar relative to atmospheric pressure, but may be quite variable.

(27) Each membrane sits on a membrane support 20, such as a plate, mesh or grate, which is arranged on the gas side of the membranes 19 within the gas loop. Below each membrane support 20 is a low height disc shaped gas collection chamber 21 on the gas side of the membrane 19. The gas loop flow initially enters a gas flow inlet 12 and passes through gas collection chamber 21 on the gas side of the membrane 19 in the second membrane unit 2b. From the gas collection chamber 21 on the gas side of the membrane 19 of the second membrane unit 2b, the gas flow passes to the gas collection chamber 21 on the gas side of the membrane 19 in the first membrane unit 2a, via a not shown gas flow connection conduit. The flow then exits the membrane block 2 via the gas flow outlet 11 and to the first 3-way magnetic valve 13, then to the second 3-way magnetic valve 14. From the second 3 way valve 14, the gas flow enters into the CO2 gas 15 H2S gas sensor (not shown but arranged instead of CO2 sensor 15). Finally the gas flow passes through to the air pump 16, that drives the gas back to the membrane block gas flow inlet 12 on the gas flow side 21 of the second membrane unit 2b.

(28) In a variant shown in FIG. 6, in which both CO2 or total carbonate as well as H2S is detected, the air loop passes from the second 3-way valve 14, the gas flow enters into the H2S detection unit 29, followed by the CO2 gas sensor 15. Finally the gas flow passes through to the air pump 16 that drives the gas back to the membrane block gas flow inlet 12 on the gas flow side 21 of the second membrane unit 2b.

(29) Otherwise the system shown in FIG. 6 is in principle identical to the system shown in FIG. 1b.

(30) The liquid sample flow path and/or the membrane block may comprise a heater element (not shown) to increase the liquid sample temperature and/or maintain constant elevated temperatures within the sample flow and/or within the membrane units.

(31) Preferably, the heating element is arranged in the liquid sample flow path 22 through the mixing block 25 or in the reaction chamber 27 so as to heat the aqueous liquid sample prior to passing the sample to the membrane block 2.

(32) The CO2 gas sensor 15 uses infrared technology as described above, so condensation does not damage the sensor as in older technologies. However, excessive condensation can lead to too high results. Condensation is managed by a combination of hydrophobic membranes 19, preventing water vapour from entering to the gas loop 10, and by regular flushing of the gas loop with atmospheric air from the outside environment using the breather valve assembly represented by the two 3-way valves 13,14. See FIG. 3. During the flushing sequence, the first and second 3 way valves 13, 14 are both activated while leaving the air pump 16 on.

(33) The H2S gas detection unit 29 detects H2S gas in the gas loop, Any commercially available H2S gas detector/sensor that is applicable for detection of H2S in gaseous environments can in principle be used.

(34) The preferred H2S detection 29 unit uses an electrochemical measuring cell. This measuring cell contains an electrolyte, a measuring electrode (anode), a counter electrode (cathode) and a reference electrode. An electric signal proportional to the pollutant combination is produced in the measuring cell. This electric signal is amplified and used for the measurement. The measuring cells use the capillary diffusion barrier technology. The use of capillary diffusion barrier technology and an additional temperature compensation avoid a negative effect caused by fluctuating air pressure and temperature.

(35) The H2S detection unit comprises a sensor unit 30 and a transmitter unit 31 (see FIG. 8). The sensor unit 30 of the H2S detection unit 29 is arranged in the gas loop. A distribution cap 32 is provided around the sensor unit 30, preferably in air tight manner. The distribution cap 32 comprises an inlet 33 and an outlet 34 to which the gas loop is connected.

(36) The H2S detection unit 29 is arranged in the gas loop to substitute the CO2 sensor 15 shown in FIG. 1 when only H2S gas is to be detected.

(37) FIG. 6 shows a possible arrangement of the gas loop where the H2S detection unit 29 is arranged in series with the CO2 sensor 15. The H2S detection unit 29 can be arranged upstreams to the CO2 sensor 15 as shown in FIG. 6, or alternatively, the H2S detection unit 29 is arranged downstreams of the CO sensor 15.

(38) Flushing the gas loop 10 using the 3-way valve combination (FIG. 3) is also important to allow the CO2 gas partial pressure in the gas loop to rapidly drop to a near zero CO2 level between measurements. This provides a rapid turn-around between measurements. This also avoids that CO2 gas present in the gas loop 10 from one sample measurement is carried over in the next sample measurement and thus provides false results.

(39) Zero CO2 gas level in the gas loop 10 corresponds to average CO2 gas concentration in atmospheric air. At present approximately atmospheric CO.sub.2 concentration has a level around 400 parts per million. Clean atmospheric air does in general not contain any traceable amounts of H2S.

(40) A near zero CO2 and/or H2S gas level between measurements is important to ensure an earlier measurement does not impact upon a future measurement. An earlier measurement may for example be very high, as compared to the next coming measurement. Without the ability to flush the gas loop of the old sample air, then the CO2 and/or H2S in the gas loop would need to diffuse across the membrane back to the water, again driven by the gas partial pressure gradient towards equilibrium as per Henry's Law. As the gradient difference is often not large, this is a slow process.

(41) During normal measurement operation, the 3-way valves combination 13, 14, are not activated as illustrated in the left part of FIG. 3. This means gas simply flows through the valves 13,14 in a closed gas loop 10 with no connection with the outside environment. In order to flush the gas loop 10 with atmospheric air, both the 3-way valves 13,14 are activated simultaneously. When activated, the first 3-way valve 13 forces air coming from the membranes out of the air loop to the outside environment as shown in the right part of FIG. 3 and as illustrated with the arrow 13′. The second 3-way valve, 14 ensures that the suction force created by the air pump sucks in new air into the gas loop as illustrated with arrow 14′ (FIG. 3).

(42) Another important function of the CO2 probe is the ability to measure the entire carbonate buffer system concentration of a fish farm (primarily free CO2 and HCO.sub.3.sup.−). To do this acid is added to the water sample to measure any bound carbonate forms. As the sensor head is designed to read only free CO2 gas, a small dose of citric acid to the water sample (to <pH 4 or to pH<3 if H2S is also to be detected see further below), drives the carbonate buffer system completely to the left, ensuring nearly the entire buffer system becomes free CO2 gas. The final measurement is effectively the concentration of CO2 and HCO3— together.

(43) FIG. 4 illustrates the sample preparation for determination also of HCO3— concentration in the sample flow or thus the total carbonate content of the liquid sample. In order to force dissolved bicarbonate and carbonate ions to CO2 in gas form, pH is lowered to below pH 4 in the sample flow by addition of an acid. The acid is transferred from a not shown acid container.

(44) When measuring the content of gaseous CO2 in the samples, the first and second bypass valves 23, 24 (FIG. 4a) remain deactivated. Thereby, the aqueous liquid sample can flow through the bypass valves 23, 24 and directly to the membrane block 2 of the CO2 detection and monitoring system 1. When the bypass valves 23, 24 are activated, the aqueous liquid sample is directed to a mixing block 25 FIG. 4b). When the bypass valves 23, 24 are activated an acid dosing pump 26 is also activated, delivering fine drops of acid, e.g. a citric acid solution, to the mixing block 25. Thereby pH of the aqueous liquid sample rapidly drops to below pH 4. The aqueous liquid sample enters a reaction chamber 27 to ensure a residence time that enables HCO3— to chemically shift from dissolved HCO3— to free CO2 gas. The acidified aqueous liquid sample is then directed to the CO2 membrane block 2 of the detection system 1.

(45) In the CO2 and/or H2S detection system, the high level of free CO2 and/or H2S enables a rapid diffusion across the membranes in the membrane block 2 and ensures rapidly obtaining the equilibrium across the membranes when detecting total carbonate content in the acidified samples.

(46) FIG. 7a describes a variant of the membrane block, where the first and second membrane units 2a, 2b are arranged in a side-by-side manner. This allows for lower overall building height, easier access to the membrane chambers, e.g. for maintenance or re-pair, and for better condensation removal characteristics.

(47) Further, it is possible using two or more membrane blocks 2 in series, providing 4 membranes (or more) for gas separation rather than 2, see FIG. 7b.

(48) Further, it is possible using two or more membrane blocks 2 in parallel, e.g. as discussed above when detecting CO2 gas with one detection and sampling unit while a second parallel detection unit receives acidified sample flow from the reaction chamber 27 so as to detect total carbonate content and/or H2S.

(49) The entire process is controlled by a programmable logic controller (PLC). The PLC or another controller may also calculate the CO2, the total carbonate and/or dissolved carbonate and/or H2S concentration for further use, e.g. as a control parameter in a fish tank or pond in a land based fish farm.

EXAMPLES

(50) Several tests were made to assess the CO.sub.2 detection system as shown in FIGS. 1-2. The same test sample was used for all tests: freshwater, 20° C.

(51) The operating sequence up to and during a measurement was identical in all examples. The new CO2 detection system as shown in FIGS. 1-2 was used in all examples. When explicitly noted one membrane unit was disabled to perform comparative examples. The system with one membrane block disabled simulates commercially available CO2 detection systems.

(52) Each of the tests in the examples were repeated in five identical test runs. In all tests an initial removal of free bubbles was performed using a closed container with an airspace in which free bubbles in the sample flow may be collected, and by extracting the sample from the bottom of the container.

(53) In all examples a water sample flow of approx. 500 mL/min was used, The pressure in the liquid sample flow in all examples was raised to 200 mbar above atmospheric pressure on the liquid side of the membranes. All examples were conducted at ambient temperature, i.e. 21-22° C.

(54) The air pressure in the air loop varied with the content of CO2 in the sample. Usually, the pressure in the air loop was only raised slightly above to atmospheric pressure during measurements and was between 20-60 mbar relative to atmospheric pressure, but may be quite variable depending on the CO2 content in the sample.

(55) A PTFE membrane having a pore size of 0.02 microns was used in both membrane units in all examples. The membrane in each membrane unit was circular with a diameter of 76 mm.

Example 1

Comparing Two Membranes Versus One Membrane, in Relation to CO2 Gas Concentration in a Fresh Water Sample

(56) The new CO2 detection system as shown in FIGS. 1-2 with 2 membranes was initially tested (Graph 1, 2 membranes). Consistently a free CO2 gas content of 7 mg/L was achieved. Where the maximum CO2 concentration was evaluated to be 7 mg/L, a T.sub.90 (time at which 90% of the total gas is measured) of about 4 minutes was possible.

(57) The new free CO2 detection system as shown in FIGS. 1-2 with 1 membrane disabled was then tested (Graph 1, 1 membrane). This set-up would correspond to commercially available CO2 detection systems having a single membrane in membrane block which isolates gaseous CO2 from the liquid sample. A free CO2 gas content of between 5 and 6 mg/L was achieved. Where the maximum CO2 concentration was evaluated to be 7 mg/L, a T.sub.90 about 7 minutes was possible.

(58) FIG. 5a shows sample measuring time on the X-axis (in minutes) and the CO2 gas content detected by the sensor and computed into CO2 concentration (in mg/l) in the sample on the Y axis. Each line represents a repeated identical test run.

(59) Test 2; 2 Membranes Versus 1 Membrane, dissolved HCO3— Concentration

(60) The new free CO2 detection system as shown in FIGS. 1-2 with 2 membranes was initially tested (Graph 2, 2 membranes). Consistently a total of HCO3— content of 190 mg/L was achieved within 4-5 minutes.

(61) The new free CO2 detection system as shown in FIGS. 1-2 with 1 membrane disabled was then tested (Graph 2, 1 membrane). A HCO3— content of a little over 80 mg/L was achieved after 12-13 minutes.

(62) FIG. 5b shows sample measuring time on the X-axis (in minutes) and the CO2 gas content detected by the sensor and computed into total carbonate/CO2 concentration (in mg/l) in the sample on the Y axis.

Example 3

Standard Diffusion Time Between Measurements

(63) The new free CO2 detection system as shown in FIGS. 1-2 with 1 membrane disabled was tested (FIG. 5c) for determining how long old sample air in the gas loop would need to diffuse across the membrane back to the sample water, driven by the gas partial pressure gradient towards equilibrium as per Henry's Law.

(64) This test was made following a HCO3— measurement where the end concentration in the gas loop was high.

(65) This test illustrated it would take about 5-6 minutes for the CO2 partial pressure in the air loop to drop to atmospheric levels, or corresponding to 0 mg/l in the sample flow and thus to be ready for a new measurement.

(66) The detection system as shown in FIGS. 1-2, including membrane block, combined with the breather valve assembly for flushing old sample air from the air loop, can be at 0 mg/L in <5 seconds (corresponding the level of CO2 in atmospheric air).

(67) FIG. 5c shows sample measuring time on the X-axis (in minutes) and the CO2 gas content detected by the sensor and computed into CO2 concentration (in mg/l) in the sample on the Y axis.

Example 4

Comparing H.SUB.2.S Sensor Measurement Against Expected Concentrations

(68) Several tests were made to assess the H.sub.2S detection system as shown in FIGS. 1-2 but with a H2S detection unit 29 as described above instead of the CO2 sensor 15.

(69) The same test sample was used for all tests: freshwater, 20° C.

(70) The operating sequence up to and during a measurement was identical in all examples and as described in examples 1-3 in relation to detection of CO2 in the setup with two membrane units in series.

(71) Each of the tests in the examples were repeated in 2 identical test runs. In all tests an initial removal of free bubbles was performed using a closed container with an airspace in which free bubbles in the sample flow may be collected, and by extracting the sample from the bottom of the container.

(72) In all examples a water sample flow of approx. 500 mL/min was used, the pressure in the liquid sample flow in all examples was raised to 200 mbar above atmospheric pressure on the liquid side of the membranes. All examples were conducted at ambient temperature, i.e. 18-20° C.

(73) A PTFE membrane having a pore size of 0.02 microns was used in both membrane units in all examples. The membrane in each membrane unit was circular with a diameter of 76 mm.

(74) A stock solution was made with 1 mg of Na.sub.2S in 20 L of distilled water. From this there was made three sample solutions with respectively 5, 10, 15 and 20 ml of stock solution mixed in 10 L of distilled water.

(75) From this the actual H.sub.2S concentration in the sample solution can be calculated, these are the target values for the experiments. They can be seen in table 1, below

(76) TABLE-US-00001 TABLE 1 Sample solution(ml) 5 10 15 20 H.sub.2S 3.54 7.00 9.57 12.77 concentration(μg/L)

(77) The experiment was performed by running the water sample past 2 membranes, letting the H.sub.2S diffuse across into the air loop and past the H.sub.2S sensor. Before the H.sub.2S in the sample could diffuse out of the water, it had to be warmed up to above 35° C. and have pH lowered to below 4. Warming the water sample was done with a heating element around the reaction chamber. Decreasing pH to below 4 was achieved by citric acid dosing.

(78) For each sample solution two duplicate experiments were performed. FIG. 6a shows the results from the experiment. There is a clear correlation between the target concentration of H2S (grey block in FIG. 8) and the H2S concentration level detected in the two parallel test runs (blue/orange blocks in FIG. 9)

REFERENCE NUMBERS

(79) 1. CO2 gas detection system 2. Membrane block a. First membrane unit b. Second membrane unit 3. Liquid sample inlet 4. Liquid sample outlet 5. Arrow indicating inlet flow 6. Arrow indicating outlet flow 7. Flow connection between first and second membrane units 8. Outlet of first membrane unit 9. Inlet on second membrane unit 10. Gas flow loop 11. Outlet of gas flow side of membrane block 12. Inlet of gas flow side of membrane block 13. First three-way valve in breather valve assembly 14. Second three-way valve in breather valve assembly 15. CO2 gas sensor (IR sensor) 16. Air pump 17. Air loop flow direction arrows 18. Liquid flow path 19. Membrane 20. Membrane support 21. Low height disc shaped chamber on gas flow side of membrane 22. Sample flow 23. First bypass valve 24. Second bypass valve 25. Mixing block 26. Acid pump 27. Reaction chamber 28. Bypass flow path for acidification of liquid sample 29. H2S detection unit 30. Sensor unit 31. transmitter unit 32. Distribution cap 33. Gas loop inlet in distribution cap 34. Gas loop outlet of distribution cap