The tendency of natural gas to form small amounts of liquid with decreasing pressure is a well-known phenomenon called retrograde condensation. The formation of liquid in the transmission grid is undesired because the presence of liquids can both result in failures of components in the transmission system itself (for example compressor facilities) as well as cause severe damage to the facilities of end-users (for example to the hot gas path components of a gas turbine). Therefore, for all parties involved in the chain from gas production to gas usage, it is important to define an unambiguous property describing the (retrograde) condensation behaviour of a natural gas for which specification can be drawn up in delivery and transport contracts. Frequently the property of hydrocarbon dew point is used for this purpose. Another less frequently used property, defined in the ISO 6570 standard, is the potential hydrocarbon liquid content (PHLC). The measuring techniques associated with these two properties differ from each other. Direct measuring methodsto determine the hydrocarbon dew point at a given pressure are based on the detection of the first droplet(s) of liquid on a mirror surface, which is continuously lowered in temperature, whereas methods to determine the PHLC are based on the gravimetrical determination of the amount of liquid formed at a given pressure and a given temperature.
The characteristic condensation behaviour of a natural gas, the so-called phase envelope, is given in Figure 1.
Typically for a treated pipeline quality natural gas, the maximum condensation temperature occurs at pressures between 20 bar and 40 bar depending on the exact composition of the natural gas, whereas for a pure substance the maximum condensation temperature occurs normally at maximum pressure. The maximum temperature independent of the pressure at which condensate is formed is called the cricondentherm temperature. The pressure at which the cricondentherm temperature occurs is called the cricondentherm pressure. Within the two phase region (-left from the dew point curve)-condensate will be formed. At a given pressure, lowering the temperature will result in the formation of more condensate.
As stated before, two different direct measuring methods are frequently used for determining the condensation behaviour of natural gases.
Chilled mirror methods, either manually or automatically operated, are commonly used to determine the hydrocarbon dew point.
Gravimetric methods based on the ISO 6570 standard[1], either manually or automaticly, to determine the potential hydrocarbon liquid content.
In practice, manual cold mirror devices are used as the "de facto" standard to which both automatically operated chilled mirror analysers and calculation methods are adjusted. Since natural gas is more and more transported over long distances, passing transmission networks of different operators, the traceability of a measurement is becoming an important issue.
Due to the absence of hydrocarbon dew point reference materials and reference instruments, it is not possible to calibrate commercially available hydrocarbon dew point analysers in a traceable way. Since the actual implementation of the measuring principle differs from instrument to instrument, analysers from different manufacturers may give different values for the hydrocarbon dew point for a given gas. In practice, the dew point of an automatic dew point monitor is often "tuned" to match the value measured by a manually chilled mirror, or "tuned" to the value calculated from the known gas composition using a thermodynamic model.
Moreover, from their working principle there are two major sources, which can be responsible for significant systematic errors in the measured hydrocarbon dew point, and which cannot be adjusted for because no proper calibration method exists.
Often a significant amount of liquid needs to be formed on the mirror, before the instrument is able to detect the dew point temperature. From calculations it can be shown that the amount of condensate, which is required to get a reproducible dew point observation often corresponds to a liquid drop-out of 20 mg/m3 (n)~50 mg/m3 (n). In fact, the dew point meter can be considered as a condensate meter, and the measured dew point temperature is actually the equilibrium temperature at a certain PHLC value, for example at 30 mg/m3 (n) liquid drop-out.
The cooling rate of the mirror of a dew point analyser appears to be another important parameter for a proper dew point measurement. Measurement of the dew point occurs in a small sized cell with the bottom surface made of polished metal. Only this bottom side is cooled, resulting in temperature gradients in both the measurement cell itself as well as in the gas inside the cell. Theoretically the applied cooling rate for the mirror surface should be so small that the gas temperature is always in equilibrium with the mirror temperature and that there is enough time for the condensate to drop out on the mirror surface. In practical applications, the cooling rate is often much higher. A high cooling rate results in a lag between the decrease in gas temperature and the condensate drop-out onto the mirror. Also, the temperature sensor, which is mounted in the vicinity of the mirror surface, will record a lower temperature than the actual surface temperature of the mirror. Both effects will result in readings at much lower temperatures than the real dew point temperature.
The traceability issue with hydrocarbon dew point measurements was the main reason for Gasunie to use the property of potential hydrocarbon liquid content to define the condensation behaviour of a natural gas. Contracts based on potential hydrocarbon liquid content contain a specification, such as having the potential hydrocarbon liquid content shall be equal to or less than 5 mg/m3 (n) at a temperature equal to or higher than -3 ℃at all pressures ranging from 1 bar to 70 bar. Given the shape of the phase envelope, the measurement of hydrocarbon dew point and potential hydrocarbon liquid is usually carried out at a pressure between 25 bar and 30 bar where liquid drop out occurs at the highest temperatures (see Figure 1).
In the 1980's and 1990's Gasunie, together with the Delft University of Technology, carried out an extensive research project[4-5], called PHLC prediction, to develop an alternative indirect method to determine the PHLC value of a natural gas based on an extended GC analysis in combination with an equation of state because of the labour intensive character of the ISO 6570 PHLC measurements. This research project showed that although it is in principle possible to predict the PHLC values of a natural gas correctly with such a method, in practice it is extremely hard to operate an unattended process analyser in such a way that the high requirements on the heavy hydrocarbon analysis are always fulfilled.
Gasunie decided to look into the measurement of hydrocarbon dew point in more detail because of two developments:
• The first development was the agreement within EASEE-gas (European Association for the Streamlining of Energy Exchange) in 2005 to harmonise the hydrocarbon dew point specification for H-type gas throughout Europe (Common Business Practice 2005-001/01) [6].
• Another development was the introduction of new generation hydrocarbon dew point analysers with more sensitive detectors and advanced temperature control algorithms which resulted in measurements with a better reproducibility. Based on the experience gathered during the PHLC prediction project, Gasunie decided not to look into hydrocarbon dew point calculation methods on the basis of a gas analysis, but to limit its investigations to automatic cold mirror devices.
• Besides the determination of the measuring performance of automatic cold mirror devices, research was also carried out in the field of traceability of hydrocarbon dew point measurements. The importance of traceability of hydrocarbon dew point measurements is also recognized by EASEE-gas. In the aforementioned CBP, it is stated that the need for introducing a harmonised measurement method has been identified.
The research project carried out by Gasunie, resulted in the development of a calibration method for hydrocarbon dew point chilled mirror devices based on potential hydrocarbon liquid content measurements as described in the ISO 6570 standard. The project clearly showed that the adjustment, and even the calibration, of the new generation hydrocarbon dew point analysers against the ISO 6570 standard is possible, resulting in a measured hydrocarbon dew point value which corresponds unambiguously to a given liquid drop-out at the measured dew point temperature. In this way a traceable measurement of the hydrocarbon dew point is possible.
Based on the outcome of the project, the Technical Committee 193 'Natural Gas' decided to lay down this calibration procedure in an ISO Technical Report. The required harmonised measuring method, which still has to be identified, can clearly benefit from the proposed traceable calibration procedure of the measured hydrocarbon dew point against the ISO 6570 standard.
In this paper, a brief overview of the working principle and the limitations of today's hydrocarbon dew point analysers, calculation methods, and gravimetrical methods will be given. A description and examples of the performance of the applied dew point analyser and the gravimetrical method will be presented and subsequently results of comparison tests for different natural gases will be given. The results will be evaluated and conclusions will be drawn with respect to obtaining harmonised hydrocarbon dew point values.
The PHLC of natural gas is defined as the measurement of the condensable liquid (in milligrams) at the pressure and temperature of the measurement per unit volume of gas at normal conditions (0 ℃, 1.013 25 bar(a)). The procedure for measuring PHLC is described in the ISO 6570 standard. This international standard states that the quantity of condensate which can be formed at a certain pressure and temperature is determined by passing a representative sample of gas through the apparatus where it is first brought to the required pressure and then cooled to the required temperature. The liquids formed during cooling are separated from the gas flow and collected by means of a cyclone separator. In the manual method, as described in the ISO 6570 standard[1], this is actually done by comparing the mass of the condensate separator at the start and the end of the measuring period.
Gasunie developed an automatic method, which is derived from the measuring system used in the manual method. The principle of this automatic method is also described in the current version of the ISO 6570 standard. The main difference between the automatic and the manual ISO 6570 method is in the weighing of the collected liquid. In the automatic PHLC method, a differential pressure transmitter indirectly determines the mass of the liquid in the measuring tube underneath the cyclone separator. A schematic overview of the Gasunie Automatic Condensate Meter (GACOM®) is shown in Figure 2.
The measuring tube is automatically drained when it is completely filled with liquid. The liquid from the measuring tube is collected in a condensate drum. This method requires calibration of the differential pressure transmitter. By dosage of known amounts of a calibration liquid (-usually n-decane) -into the measuring tube at the temperature and pressure at which the PHLC measurement will be performed, the differential pressure transmitter is calibrated. The pressure and temperature sensors are calibrated once a year and the gas flow is measured by a thermal mass flow meter, which is checked at regular intervals against a calibrated wet gas meter. In this way, reliable and accurate measurements can be performed.
The uncertainty in PHLC-value depends on the pressure and temperature set points and the liquid drop-out behaviour of the gas to be measured. Values above 5 mg/m3 (n) are detected unambiguously. At low PHLC-values (< 30 mg/m3 (n)), the uncertainty is < 5 mg/m3 (n).
Because all sensors applied in the GACOM are calibrated at a regular interval against standards, it is ensured that the measurement conditions are traceable and that two distinct GACOM units will give the same results.
The automatic condensate meter is operated with a gas flow of 1 m3/h (n) at a fixed pressure as close as possible to the cricondentherm pressure (most often between 27 bar and 30 bar) and a fixed temperature (e.g. -3 ℃) at which the hydrocarbon liquid content (often 5 mg/m3 (n)) is specified in contracts. A new value for the liquid drop-out content is reported every 30 minutes. Based on extensive experiments carried out in the past, the reproducibility is almost always within ± 5 mg/m3 (n). Later on the results of an experiment with stable gas composition will be shown, from which it can be concluded that the random error (2σ) is ± 2 mg/m3 (n).
As an example, the results of two individual indirect automatic weighing instruments are presented in Figure 3. Both devices measure the potential hydrocarbon liquid content at the same operating conditions for the same gas.
Theoretically, the hydrocarbon dew point is the temperature at which the first small droplets of liquid are formed at a fixed pressure. In practice, all dew point measurement methods are based on the observation of the formation of a film of hydrocarbon condensate on the surface of an illuminated cooled mirror. The observation can be done visually (manual mirror) or by an electronic sensor (automatic chilled mirror). It is necessary to realize beforehand what a chilled mirror dew point monitor is actually determining; certainly not the true thermodynamic dew point, but a temperature corresponding to a predetermined threshold value of the detector signal (automatic mirror device) or the sensitivity of the eye of the observer (manual mirror device).
Although there are some significant differences in the implementation, the measuring principle of an automatic hydrocarbon-dew-point chilled-mirror instrument is identical for all instruments. After pressure reduction, the gas is passed through a measuring cell. During normal operation, the measuring pressure is chosen to be close to the cricondentherm pressure (the value at which the dew-point temperature is at its maximum). The measuring cell has an observation window at one side and a mirror surface at the other side. This mirror is mounted on a cooling body and the cooling down can be accomplished in a controlled way. The cooling body itself is cooled either electrically (Peltier element) or by the expansion of carbon dioxide or another gas. The temperature of the mirror is measured continuously. The sample gas may be flowed through the cell continuously, or, having flowed sufficiently to purge the cell, mirror and pipe work, blocked in without flow while the cooling cycle of the mirror starts. The mirror surface is observed by reflected light, either visually by an operator in the manual version or by photocell in the automated instrument.
An example of an implementation of an automatic dew point chilled-mirror instrument is shown in Figure 4.
The development of the calibration procedure by Gasunie as described in the paper was carried out using a Condumax Ⅱ instrument of Michell Instruments Ltd. The Condumax Ⅱ fulfils the general description in the previous paragraph. In more detail, the measurement cycle of the Condumax Ⅱ consists of two phases: the measurement phase followed by a recovery phase. At the start of the measurement phase a solenoid stops the gas flow through the sensor cell. Subsequently, the mirror is cooled down in a controlled way. The condensation of heavy hydrocarbons on the mirror surface results in an increase of detector signal. An example of such a response curve can be seen in Figure 5. The mirror temperature decreases until the detector signal matches with a factory set threshold (trip point) value. The value of this threshold is determined by the manufacturer but can be changed by the end-user. The mirror temperature, for example T1, at the threshold value TP1 can be presented as the hydrocarbon dew point measurement result. However, by setting the threshold value TP2, the mirror temperature T2 will be presented as the hydrocarbon dew point. The construction of the measurement cell is also shown in Figure 5. The cooling rate is optimised after every measurement cycle. At the start of each cycle the mirror is cooled down fast (1 ℃/s). Upon approaching the dew point temperature the cooling rate gradually decreases down to 0.02 ℃/s. In this way the measurement cycle time can be kept relatively short without the disadvantages of a high cooling rate near the dew point temperature, which would result in a significantly lower measured hydrocarbon dew point value.
After the measurement phase the mirror is heated to 50 ℃ for at least 5 minutes and the gas flow through the sensor starts again. During this recovery phase all heavy hydrocarbons condensed on the mirror will re-evaporate. In this way contamination of the mirror surface with heavy hydrocarbons is prevented.
The new calibration procedure is based on the direct relationship between hydrocarbon dew point and the potential hydrocarbon liquid content. The hydrocarbon dew point chilled mirror device is thus calibrated against the ISO 6570 standard. As a result, the measured hydrocarbon dew points by the calibrated cold mirror device will correspond to a given potential hydrocarbon liquid content. This amount, the PHLC reference value, needs to be specified by the user as an input for the calibration procedure. Subsequently, the requirements the chilled mirror device needs to fulfil to be used with the proposed calibration procedure, and the validity of the calibration for various types of natural gas and some remarks with regards to the PHLC reference value will be discussed. First, some details of how the outputted hydrocarbon dew point value of a chilled mirror device can be altered, will be given.
Depending on the instrument, the necessary adjustment after a calibration can be made in two ways:
• Changing the threshold value; by using a cooling curve (graph of the detector signal versus mirror temperature) the required detector signal is determined at which the mirror temperature equals the measuring temperature yielding the PHLC reference value. In general, this method is favoured because this is the correct method from a physical point of view; as a result of using a lower threshold value the instrument will detect the hydrocarbon dew point value with less liquid formed on the mirror surface.
• Determining an off-set value to be added to the raw hydrocarbon dew point measurement value. Although this method doesn't reflect the physical processes in the instrument, this method delivers suitable results in those cases where a large temperature shift of more than 3 ℃ is observed. Changing the threshold value in such extreme cases, results in a large increase of the measurement uncertainty due to nature of the cooling curve. As can be clearly seen in Figure 5, the detector signal only varies marginally with temperature at higher temperatures.
From the previous paragraph, one could get the impression that only a limited amount of instruments can benefit from the proposed calibration technique. This is certainly not the case; the proposed calibration procedure does not require a very sophisticated automatic hydrocarbon dew point instrument. The basic requirements the instrument needs to fulfil are:
• The values of the measured hydrocarbon dew point temperature, the measuring pressure and the date/time stamp needs to be digitally available.
• The possibility to change the detection point criteria (preferably the threshold value but alternatively an off-set value which can be added to the raw measured hydrocarbon dew point value).
• The measuring method needs to be stored in non-volatile memory to safeguard the use of the correct measuring method after a power interruption.
• The repeatability at steady state conditions should be within 1 ℃ (2σ value).
• The repeatability is best determined at a constant gas composition. However, if the gas composition is frequently changing, a plot of hydrocarbon dew point versus the potential hydrocarbon liquid content may also result in a good insight in the repeatability as can be seen in Figure 8.
Gasunie has performed tests with both L-type and H-type natural gases. Within one type, there is no need to change the calibration settings of the hydrocarbon dew point instrument; once calibrated with a certain L-gas, the calibrated instrument can be used without recalibration for different L-gases. However, an instrument with optimum calibration settings for L-type gases shows an off-set when operated with an H-type gas and vice versa. Although the thermal and mass transfer processes that take place during the cooling of the mirror surface are of a complex nature and not completely understood, the hypothesis is that the calorific value of the gas is not the important parameter, but the condensation behaviour of the gas itself. In a study[3] carried out under auspices of the National Physical Laboratory, synthetic mixtures of natural gases are characterised into three classes based on their condensation behaviour:
• "low" condensation rate; for this type of gases, a decrease in temperature of 1 ℃results in the formation hydrocarbon liquids of less than 100 mg/m3 (n). In the aforementioned study, an example of a gas is described in which a decrease of 5 ℃ results in an increase of the potential hydrocarbon liquid content of only 25 mg/m3 (n).
• "mid" condensation rate; for this type of gases, a decrease in temperature of 1 ℃ results in the formation of hydrocarbon liquids in the range between 100 mg/m3 (n) and 300 mg/m3 (n).
• "high" condensation rate: for this type of gases, a decrease in temperature of 1 ℃ results in the formation of hydrocarbon liquids over 300 mg/m3 (n). In the aforementioned study, an example of a gas is described in which a decrease of only 0.25 ℃results in an increase of the potential hydrocarbon liquid content of 250 mg/m3 (n).
As stated before, current hydrocarbon dew point analysers measure the "dew point" temperature at a liquid drop-out ranging between 20 mg/m3 (n)~50 mg/m3 (n). A reduction of these amounts results in measured values closer to the "true" hydrocarbon dew point. To minimise the difference between the actual measurement value and the "true" dew point, the dew point analyser should determine a dew point temperature corresponding with approximately 5 mg/m3 (n) liquid drop out, being the value that can be determined accurately by ISO 6570 equipment. A further reduction of the threshold value below 5 mg/m3 (n) seems to be not viable given the detection principle used in current hydrocarbon dew point analysers. Depending on the gas composition, the difference in temperature corresponding with respectively with 30 mg/m3 (n) and <5 mg/m3 (n) can be less than 1 ℃, however the difference can also be more than 5 ℃.
Since some, especially older specifications are based on the use of a manual hydrocarbon dew point chilled mirror, which results in an higher threshold value of approximately 70 mg/m3 (n), and not all hydrocarbon dew point analysers are able to operate at a threshold value of 5 mg/m3 (n), the concept of PHLC reference value is introduced in the ISO Technical Report. This PHLC reference value is an important parameter on which it is necessary to agree before the actual calibration of an automatic hydrocarbon-dew-point chilled-mirror instrument can be carried out. As stated before, the detection limit of the automatic weighing method stated in ISO 6570 is 5 mg/m3 (n), whereas the amount of liquid which it is necessary to condense onto a mirror surface to register a hydrocarbon dew point varies between 5 mg/m3 (n) and 70 mg/m3 (n) depending on the design of the instrument. Depending on the sensitivity of the chilled-mirror detector system, it is possible to carry out a calibration at an arbitrary level starting at 5 mg/m3 (n) and upwards. Since the setting of the PHLC reference value determines the measuring behaviour of the hydrocarbon-dew-point chilled-mirror instrument, it is important to specify the PHLC reference value for use during the calibration procedure and to clearly report the PHLC reference value on the calibration report. Although it is possible to choose an arbitrary level for the PHLC reference value, it is advised in the ISO technical report to limit the choice for the PHLC reference value to the following three levels:
• 5 mg/m3 (n), being the most sensitive value, corresponding to the highest value for the hydrocarbon dew point; using this PHLC reference value, the measured values of the calibrated chilled-mirror instrument fit perfectly into the regular contract specifications based on potential hydrocarbon liquid content.
• 70 mg/m3 (n), being the least sensitive value, corresponding to the lowest value for the hydrocarbon dew point; using this PHLC reference value, the measured values of the calibrated chilled-mirror instrument correspond to the measuring behaviour of a manual chilled mirror.
• 30 mg/m3 (n), being a value intermediate between the aforementioned minimum and maximum values; in general, it has been shown that automatic chilled-mirror instruments are capable of operating reliably and with low measurement uncertainty at a level of 30 mg/m3 to 40 mg/m3.
During the field tests the Condumax Ⅱ is installed in a GACOM unit. The GACOM is always connected to a high pressure gas sampling probe, which samples the natural gas from the middle of the pipe line, thus avoiding interference by liquids adsorbed or running along the pipe wall. Gas at line pressure (50 bar~70 bar) flows through a traced stainless steel tubing (< 10 meter) to the GACOM unit. In the GACOM unit itself, the tubing is traced as well. Before the gas enters the Condumax Ⅱ, it passes a membrane filter and a heat traced pressure regulator (set at a fixed pressure in the range between 27 bar and 30 bar). With a needle valve the gas flow through the Condumax Ⅱ is adjusted to approximately 30 L/h (n). A photo of the measuring set-up within the cabin is given in Figure 6.
The first results obtained with L-gas are shown in Figure 7. The GACOM was operated at 27.3 bar and a temperature of -3.3 ℃. The Potential Hydrocarbon Liquid Content-value (that is the amount of condensate formed at these conditions) varied between 40mg/m3 (n) and 200 mg/m3 (n). The variation in liquid drop-out is caused by changes in the gas composition, particularly small changes in the heavy hydrocarbon concentrations. The Condumax Ⅱ was operated at exactly the same pressure and with the standard factory settings with respect to the detector threshold value. The measured hydrocarbon dew point varied between -3℃ and 1 ℃. The grey coloured bands in Figure 7 clearly show the good relationship between PHLC-value and hydrocarbon dew point. An increase in dew point corresponds with an increase in PHLC-value and vice versa.
The relationship between PHLC-value and hydrocarbon dew point is graphically shown in Figure 8. From this figure it can be concluded that dew point and PHLC-value are directly related to each other and can both be used to monitor the 'condensation behaviour' of the natural gas.
It is evident that small changes in dew point correspond to relatively large changes in PHLC-value. From Figure 8 it can be concluded that for this particular gas, a dew point variation of 1 ℃ corresponds with a variation in PHLC of approximately 30 mg/m3 (n).
The amount of condensate, which is required to detect the hydrocarbon dew point, can be found quite easily. When the hydrocarbon dew point temperature equals the temperature of the cooling bath of the GACOM, the amount of liquid drop-out formed in the GACOM corresponds with the measured dew point. From Figure 7, this situation occurs already at the second day of the measurement. The first grey band shows hydrocarbon dew point values around -3.2 ℃and the corresponding PHLC-values around 40 mg/m3 (n)~45 mg/m3 (n) at -3.3 ℃.So, for this L-gas, a liquid drop-out ofapproximately 40 mg/m3 (n) is required to detect the hydrocarbon dew point using the factory setting for the threshold value. This experimental observation supports the conclusion drawn in the previous paragraph that dew point analysers are detecting the "true" dew point, but an equilibrium temperature at which a significant amount of condensate is already formed on the mirror surface.
To verify the amount of condensate formed on the mirror during stable gas conditions, the bath temperature of the GACOM was adjusted to the hydrocarbon dew point value measured by the Condumax Ⅱ. The results are shown in Figure 9. Again these results confirm that the measured hydrocarbon dew point value corresponds with approximately 40 mg/m3 (n) at the same pressure and temperature.
From Figure 9 it can be concluded that, over a period of more than 12 hours, the dew point remains at a value of -0.7 ℃, only varying within ±0.1 ℃and the PHLC-value of 40 mg/m3 (n) with maximum variations of ±2 mg/m3 (n). These results confirm the stability and the small random errors in the measured values of both instruments.
It can be concluded that, by applying the factory settings of the dew point analyser, the obtained hydrocarbon dew point value and the equilibrium temperature at a PHLC-value of 5 mg/m3 (n) according the ISO 6570 are not in close agreement and that the measured hydrocarbon dew point is surprisingly lower compared to this PHLC equilibrium temperature. However, the direct relationship between PHLC-value and dew point allows the change of the threshold value of the hydrocarbon dew point analyser, so that the measured dew point will correspond with a significant lower PHLC-value of approximately 10 mg/m3 (n).
The detector signal of the Condumax Ⅱ does increase with decreasing mirror temperature due to the formation of more and more liquid on the mirror surface as can be seen in Figure 5. Therefore, lowering the threshold value will require less liquid formation on the mirror, before the dew point temperature is 'detected'. This 'detected' dew point value will be closer to the "true" hydrocarbon dew point of the gas.
Using the data of previously obtained cooling curves, it can be estimated that to 'match' the measured "dew point" with the equilibrium temperature at a value of 10 mg/m3 (n) liquid drop-out, the value of the dew point measured by the Condumax Ⅱ should be increased by approximately 2 ℃ for these L-gases. This temperature shift can be introduced in the Condumax Ⅱ outcome by a decrease in threshold (trip point) value from 275 mV to 165 mV. To verify this adjustment in practice, the threshold value was decreased accordingly and again measurements were carried out with the same L-gas. The results for the measured dew points and PHLC-values measured at the same pressure and a cooling bath temperature of -2 ℃ are shown in Figure 10.
During the measurements the gas quality frequently changed and resulted in large variations in both hydrocarbon dew point and PHLC. The PHLC-value varied between 0 and 60 mg/m3 (n) and the hydrocarbon dew point between -5 ℃ and 2 ℃. The reduced threshold (trip point) value of 165 mV still results in a good relationship between hydrocarbon dew point and PHLC-value. The amount of condensate, which is required to detect the hydrocarbon dew point, can be read from Figure 10; the first and third grey coloured band show that a hydrocarbon dew point of -2 ℃ corresponds with approximately 10 mg/m3 (n) liquid drop-out, formed at an equal bath temperature of -2 ℃. So, it can be concluded that a change in the threshold value of the detector signal, indeed results in an improved relationship between the measured hydrocarbon dew point and a low PHLC value (10 mg/m3(n)) for the potential hydrocarbon liquid content without deteriorating the outcome of the instrument.
Since more and more natural gas is being shipped over longer distances through networks of various operators, the necessity for a traceable hydrocarbon dew point measurement is rising.
The good relationship between PHLC-value and measured dew point is used as basis for the development of a traceable on-site calibration procedure for hydrocarbon dew point meters. This procedure, which is now described in a ISO technical report ISO/TR 12148, is based on the ISO 6570 standard.
Experimental work carried out by Gasunie on a commercially available automatic hydrocarbon dew point instrument, showed that in the daily practice the proposed calibration procedure is also applicable.
Although the calibration procedure is probably valid within a family of natural gases with the same condensation behaviour, additional work needs to be carried out to investigate the limitations and to define more general rules for the applicability of a calibration with a specific gas composition for gases with different composition / condensation behaviour.
2012, Vol. 41 Issue (3): 253-263
2012, Vol. 41 Issue (3): 253-263