GAS 2011 Analytical Devices for the Measure of Water Vapor in the Natural Gas Process and Transmission Industry

7/30/2019 GAS 2011 Analytical Devices for the Measure of Water Vapor in the Natural Gas Process and Transmission Industry

1/8

PROCESS INSTRUMENTS

MOISTURE

Technical

Paper

www.ametekpi.com

Analytical Devices for the Measurement of

Water Vapor in the Natural Gas Process

and Transmission Industry

D.R. Potter

AMETEK Process Instruments

2876 Sunridge Way N.E.

Calgary, AB

T1Y 7H9

PRESENTED AT:

GAS2011


2/8

9-11 February 2011 Beurs-WTC Rotterdam

ANALYTICAL DEVICES FOR THE MEASUREMENT OF WATER VAPOR IN THE NATURAL GASPROCESS AND TRANSMISSION INDUSTRY

D.R. PotterAMETEK Process Instruments

2876 Sunridge Way NE, Calgary Alberta [email protected]

+1-905-634-4401

ABSTRACT

The determination of water vapor (and water dew point) is crucial in the processing, custody transfer andtransport of natural gas. High levels of water vapor in a natural gas stream can lead to a number of problemswhich include the formation of hydrates and the contribution to freezing and corrosion of plant and equipment.There are a number of ways of determining the water vapor concentration in natural gas, including the use ofadvanced, near-infrared tuneable diode laser absorption spectrometers (TDLAS)

.

This paper will discuss AMETEKs experience with water vapor measurements in natural gas applications andhighlight the development of their laser-based instruments for the measurement of water vapor in stationary andportable applications in the natural gas and LNG industry.

1 NATURAL GAS AND THE WATERCONNECTION

Natural gas is traditionally recovered fromporous or fractured sub surface layers of manyvarious geologic formations on Earth. The recoveryprocess of natural gas generally entails sub surfacedrilling to extract the gas from these deposits(wells). Although the composition of natural gasvaries from well to well, the hydrocarbon content ofnatural gas consists primarily of light, aliphatic

hydrocarbon components comprised of methane,ethane, propane and butanes. The balance includeslower concentrations of heavier hydrocarbons andvarious gaseous non-hydrocarbons such asnitrogen, carbon dioxide, hydrogen sulfide, helium,elemental sulfur and mercury.

Natural gas at the well head is typicallysaturated with water under natural conditions, andproducers generally remove any condensibles at thewellhead to reduce the likelihood of liquid drop outin the pipeline system responsible for transportingthe gas to a processing facility for further treatment.

1.1 The Removal of Impurities in Natural GasNatural gas generally requires several

processing steps to remove or significantly reduce

the non-hydrocarbon constituents present in thegas. Of particular importance to the producers andtransporters of natural gas is the removal of the acidgas components, which consist primarily of carbondioxide (CO2) and hydrogen sulfide (H2S). Theremoval of the acid gas components reduces thecorrosivity and toxicity of the gas, which contributesto consistent quality of the final product. While it isnot in the scope of this paper to discuss the removalof all potential impurities present in natural gas, theremoval of the acid gas components warrantsparticular mention, as the processes generallyemployed to remove these impurities can re-

introduce high concentrations of water vapor intothe process stream.

1.2. Water Content in Natural GasWater, many times in combination with other

impurities and the hydrocarbons present in the gasstream, has the potential to severely damage anatural gas processing and delivery system. Sincethe delivery system in the natural gas industryconsists primarily of high pressure pipeline, anypotential for damage to the delivery system must be

minimized. In addition to reducing the water contentin natural gas pipeline transmission systems,cryogenic hydrocarbon processing requires aneven further reduction of the water content ofnatural gas.

A major source of trouble in natural gaspipelines is the formation of a gas hydrate. Ahydrate is a physical combination of water and othersmall molecules to produce a solid which has an icelike appearance but possesses a different structurethan ice [1]. Figures 1 shows a nearly fullyobstructed pipeline as a result of hydrate formation.

Figure1: A Natural Gas Hydrate

In addition to being a major contributor to theformation of hydrates, water reduces the overall


3/8


combustion quality of the gas, further increasing therequirement to remove as much of it as possible.And finally, free water in combination with any of the

acid gases present in natural gas contributes tocorrosion and cracking of materials typically used innatural gas processing and transmission lines.

2. WATER MEASUREMENT TECHNOLOGIES

For measurement of water content, water vaporor water dew point in natural gas, there aregenerally two classes of measurement devices: theprimary measurement devices and those thatmeasure concentrations. The measurementrequirement and customer choice generallydetermines which class of measurement device ischosen. Table 1 highlights the various water contentor dew point specifications found in selectedcountries.

Table 1: Typical Water Dew Point/ContentSpecifications

Analyzers capable of measuring the dew-pointtemperature directly are based on chilled mirrortechnology and are sometimes referred to as firstprinciple or primary devices. Expressing water

content in terms of dew-point originates from thisfirst principle device tracing its history back to theearly dates of the natural gas industry when theU.S. Bureau of Mines developed the manual chilledmirror instrument. Chilled mirror devices are theonly instruments to actually measure the dew-pointtemperature of a sample [2].

The second class of analyzers measuresconcentration (e.g., ppmv, ppmw, lbs/mmft

3, and

mg/Nm3) or partial pressure. For measuring water

content the technologies commonly used in thenatural gas industry include: quartz-crystalmicrobalance (QCM), electrolytic cells, capacitance-based sensors (aluminum oxide or thin-filmtechnology), fibre optic hygrometry and more

recently the development of near-infrared (IR)spectrometers. All of these techniques provide anindirect measure of the sample dew-point, which iscalculated from compositional information.

While it is well beyond the scope of this paper toprovide a detailed description of the varioustechnologies capable of determining the watercontent or dew point temperature in natural gas, it isworth highlighting the general moisturemeasurement capabilities and accuracyspecficiations provided by the various technologies.

The performance specifications of the mostcommonly used techniques is provided in Table 2.

Table 2: Accuracy Specifications of VariousMoisture Measurement Devices

For devices which measure water vapor orconcentration, the accepted method of conversionbetween concentration and dew-point is based onempirical data from the original research of theInstitute of Gas Technology (IGT). This work waspublished in the 1950s in IGTs Bulletin no. 8. Thiswork is republished in the American Society forTesting and Materials (ASTM) standard D 1142(latest revision) [2]. While this work is the acceptedstandard in the natural gas industry in NorthAmerica and many other regions, it has somesignificant limitations. Chief among these is thelower-temperature limit of 40C; use of this methodbelow 40C is an extrapolation. More recently, aEuropean calculation adopted by the European GasResearch Group (GERG) and published by the

International Organization for StandardizationSociety (ISO) as method ISO 18453:2004 specifiesa method to provide users with a reliablemathematical relationship between water contentand water dew point in natural gas, taking intoaccount the restrictions or limitations of previouslypublished data [3].

A second issue is that these data correspond todew-point values, and many instrumentmanufacturers have designed their products toreport the frost-point. While this second point is nota problem, it does add to the level of confusionregarding dew/frost-point measurements. Forconvenience, most electronic instruments includeprovisions for displaying their readings in both

dew/frost-point and concentration. However, itmust be remembered that these instrumentsare converting between the compositional(concentration) and physical (dew/frost-point)information.

3. SPECTROSCOPIC TECHNIQUES FOR THEMEASUREMENT OF WATER VAPOR INNATURAL GAS

The measurement of water vapor using thisoptical (or spectroscopic) technique has receivedvery favorable response from the natural gas


4/8


precess and transmission industry. The use of non-contact spectroscopic methods used for determiningthe water vapor concentration offer several key

benefits, chief among them being that non-contactsensors do not suffer from fouling, depletion orcontamination which plagues the chemical sensorstraditionally used for the measurement of watervapor in industrial process applications.

3.1 Tunable Diode Laser-Based TechnologyThe use of near-infrared based Tuneable Diode

Laser Absorption Spectroscopy (TDLAS) for theidentification and quantification of water vapor innatural gas process and transmission applicationshas been widely accepted.

TDLAS-based spectrometers offer severalimportant attributes:

- High specificity for the analyte (component)of interest

- High sensitivity- Fast response speed- Multi-component capability (e.g. CO2/CH4)

To further expand on the above mentionedpoints, the emission bandwidths of tunable diodelasers are on the order of 10 -4 to 10-5 cm-1 , givingthe TDLAS technique extremely high spectralresolution, which results in the ability to isolate asingle rotational-vibrational line of the analytespecies [4]. Figure 2 provides an illustration of thetypical bandwidth of a conventional IR sourceversus the TDLAS. The ability of the TDLAS toproduce a near monochromatic light sourceenables it to provide high specificity and spectral

resolution for the component of interest.

Figure 2: Conventional IR and TDLAS Bandwidth[6]

In addition to providing a fast speed of response,diode lasers operate at room temperature, are verysmall, operate at very low power (making thedevices suitable for operation as stationary andportable instruments) and have a long operationallife. Careful selection of additional diode lasers mayallow a single TDLAS platform to perform multi-component capabiity, of specific interest in thenatural gas industry are the measurement of CO2,H2S, and methane (CH4). Figure 3 shows aschematic diagram of a dual TDLAS instrument

capable of providing water vapor and carbondioxide measurements, configured with dual lasers,dual reference cells and single cell design.

Figure 3: Dual Laser TDLAS Analyzer with SingleCell, Dual Lasers and Dual Reference Cells [9]

3.2 Wavelength Modulation Spectroscopy (WMS)In addition to providing superior spectral

resolution, TDLAS spectrometers typically employenhanced detection techniques, known asWavelength Modulation Spectroscopy (WMS). InWMS, the wavelength of the optical source (thediode laser) is modulated at very high frequencies,so that a small region of the absorption spectrum ofthe analyte (e.g. water molecule), containing adistinct absorption peak, is repetitively scanned. Asthe wavelength of the source scans through thespectral feature of interest the relative attenuation ofthe source power is varied. This produces a time-

varying signal at the detector, and coupled withphase sensitive detection techniques produces anoutput signal representing the presence andquantity of the analyte of interest [4].

This type of modulation spectroscopy provides azero background technique and can provide a 10 100 times increase in sensitivity over conventionalabsorption spectroscopy technologies.

3.3. Making Water Vapor Measurement in aHydrocarbon Background

As described in the previous section, diodelasers have the ability to effectively isolate a singleabsorption line of the component of interest.However, making relatively low concentrationmeasurements of water vapor in the background of

high concentrations of primarily methane in naturalgas presents a challenge. Specifically, methane willcontribute a small amount of backgroundinterference to the measurement which, ifuncorrected, will produce a bias in the measurement[4]. Figure 4 shows recorded water vapor spectra inthe presence of bulk natural gas components in thebackground.

ELectronicsUnit

Temperature

Control

Current

Control

MirrorGRIN Lens

Photodiode

GRIN Lens

Sample cell

H2O Reference Cell

Photodiode

GRIN Lens

1854 nmLaser Diode

2004 nmLaser Diode

ELectronicsUnit

Temperature

Control

Current

Control

MirrorGRIN Lens

Photodiode

Beam Splitter

Sample cell

ELectronicsUnit

Temperature

Control

CurrentControl

MirrorGRIN Lens

Photodiode

Beam Splitter

Sample cell

1854 nmLaser Diode

2004 nmLaser Diode

ELectronicsUnit

Temperature

Control

CurrentControl

1854 nmLaser Diode

2004 nmLaser Diode

ElectronicsUnit

Temperature

Control

CurrentControl

1854 nmLaser Diode

2004 nmLaser Diode

Photodiode

CO2 Reference Cell


5/8


Figure 4: Absorption Spectra of Water Vapor (1%),methane, ethane, propane (100%)

To eliminate or reduce as much (or most) of thepotential background interference, wavelengths arechosen which provide the least amount ofinterference from the background gas.

In addition to the careful selection of theappropriate wavelength, enhanced digital signalprocessing capability available with the AMETEKProcess Instruments advanced line of TDLASanalyzers provide the capability of recording thespectrum around the selected water vapor peak,enabling an accurate measurement andcompensation of the methane present in thebackground gas. This type of design providesdistinct performance enhancement over traditionalanalog implementation of TDLAS technology, aswell as providing for the additional measurement ofmethane with a single laser.

It should also be noted at this point thatregardless of the careful selection of wavelengthand enhanced detection capability, TDLASanalyzers have minimum full scale range limitationsin natural gas backgrounds.

This is of particular importance when dealingwith natural gas processing and LNG recoveryoperations, where the nominal water vaporconcentrations typically measured at the outlet ofmolecular sieve driers are below 1 ppm. Sensor-based technologies, such as Quartz-Crystal-Microbalance (QCM) devices provide superioraccuracy and lower limits of detection currently notachievable using traditional near-infrared TDLAStechnology.

3.4 TDLAS Performance Issues

While it appears that TDLAS technology is a

nearly perfect solution for the measurement of

water vapor in natural gas, there are limitations with

the technology which should be mentioned.

TDLAS devices operate in the near-infrared region

of the electromagnetic spectrum (EMS), in the band

from 700 2500 nm. In this region, it is the weaker,

less sensitive overtone and combination absorption

bands which are found, making it inherently more

challenging to provide trace level detection of the

analytes of interest. Figure 5 provides the HiTran

spectrum of water across the near and mid- IR

region of the EMS [5] (the collection of peaks in thefar left region on the horizontal axis show the

weaker overtone absorption peaks of water found in

the near-infrared region, while the larger peaks in

the center of the horizontal axis represent primary

absorptions found in the mid-IR region of the EMS).

1 2 3 4 5 6 7 8 9 100

0.5

1

1.5

2

2.5

x10-19

Wavelength (microns)

CrossSection(cm2)

Figure 5: HiTran Spectrum of Water Vapor (1 10

micron region)

While it is possible to enhance thesensitivity of TDLAS devices by employingmulti-pass cells which effectively lengthen theoptical path of the instruments, relatively weakabsorption bands coupled with backgroundinterference (from methane present in the

natural gas) limits the effective full scale rangeof TDLAS devices used in this service.

Another major (but little discussed)negative aspect of measuring low levels orweakly absorbing compounds with TDLAS isthe difficulty of ensuring the TDLAS system isadequately and continuously focused on thevery narrow energy band. The junctiontemperature is critical to the measurement, aseven small changes in temperature will shiftthe center wavelength of the emission and canresult in erroneous measurements caused byloss of line lock [9]Figure 6 shows the spectral shift of the

fundamental wavelength as a result of atemperature change.


6/8


Spectral Index

0 50 100 150 200

2FSignal

-0.01

0.00

0.01

0.02

0.03

27.4C26.8C

25.5C

25.55C

25.54C

Spectral Index

0 50 100 150 200 250

2F

Signal

-0.08

-0.04

0.00

0.04

0.08

0.12

Figure 6: Wavelength Shift with Temperature

To overcome temperature drift issuesassociated with TDLAS devices, referencecells filled with a known concentration of therequired analyte(s), as provided with theAMETEK series 5100 TDLAS analyzers, arescanned repeatedly to ensure the laser knowswhere to focus and minimize loss of line lock ofthe instruments.Figure 7 shows an AMETEK TDLAS analyzerwith incorporated sealed reference cellassembly.

Figure 7: Integrated TDLAS Analyzer withReference Cell

4. PERFORMANCE VALIDATION ANDVERIFICATION

It is an accepted requirement that the

performance of an analytical device, fixed or

portable, needs to be validated. The measurements

produced by the instrument need to be tested

against samples that are traceable to recognized

standards (e.g., the National Institute of Standards

and Technology, or NIST). This can be

accomplished by a number of methods: bottled

calibration gases, portable permeation calibrators,

onboard permeation calibrators, sealed reference

cells with a known concentration of analyte in an

inert background gas and factory calibration by

returning the device to the manufacturer for certifiedcalibration [2].

AMETEK Process Instruments has successfullyused on board permeation devices as moisturecalibration sources for nearly four decades.Calibration sources provide the ability to verify and,if required, calibrate contact sensor devices such asthe AMETEK quartz crystal microbalance (QCM)line of products. Calibration sources peridociallyprovide the means to verify the operation of thesensor, and make necessary adjustment if drift orsensor degradation has occurred. On boardmoisture calibration devices are calibrated andtraceable to NIST.

In the case of TDLAS based systems, neitherthe laser source or the detector element come indirect contact with the process and, therefore, thereis no change in the system response relative to thesensor contamination issues described above [4].

However, as described in the previous section, itis still possible for the TDLAS analyzers to provideerroneous results. Users of process analyzersystems expect to periodically check or verifiysystem performance to ensure that the analyticalsystems are performing properly.

The use of sealed reference cells allows theuser to periodically check and verify theperformance of the analyzer system. Figure 8shows a TDLAS-based system which provides anon-board sealed reference cell.

Figure 8: TDLAS Design With Sealed ReferenceCell

A small portion of the laser output is split out andpasses through the reference cell. Data isessentially collected simultaneously from both thenatural gas stream and the water reference sampleproviding a real-time confirmation that the laser is


7/8


locked on the moisture absorption line. The waterreference cell is also used to perform a reliabilitycheck on the quantitative measurement of the water

measured in the sample cell. [4]. Verification ofwater concentration is achieved by measuring thetemperature of the reference cell and using knowncell pressure to calculate the water content in thereference cell using well established thermodynamiccalculations. The analyzer can be programmed toprovide an alarm condition if the measuredconcentration of the reference cell falls outside ofestablished thresholds of the calculated referencecell concentration.5. SAMPLE SYSTEM COMPONENTS FORWATER VAPOR DEVICES

No technical paper on water vapor analysis iscomplete without a brief discussion on samplesystem design for water vapor analyzers used in thenatural gas processing and transmission industry.

5.1,. The Water Molecule And Sample SystemSurface Interaction

Water is a small polar molecule that will stick tomost surfaces. The adsorption of water ontosurfaces in a sample system can cause asubstantial decrease in the response speed of asample system. The lower the moistureconcentration that is to be measured, the moreserious the role of adsorption becomes. When awet challenge is introduced into a dry sample line,water molecules from the gas phase will populate al lof the available surfaces. Those water moleculesthat are lost to the surface of the sample system are

not measured by the analyzer. Conversely, when adry gas is introduced to a wet sample line, thesurfaces will desorb water until equilibrium isreached. Thus, surfaces are responsible forsubstantial lag times in measuring both increasesand decreases in the moisture concentration of thesample [2].

5.2. Sample System Material SelectionALL wetted components in the sample system

(those components which come into direct contactwith the sample gas) should be chosen to preventexcessive surface oxidation. Examples of materialswhich should NOT be used include copper,aluminum and carbon steel. Any material whichpotentially creates an oxide film and promotes the

adsorption and desorption of water molecules is tobe avoided. This includes the use of any plastics orteflon.

Acceptable materials of construction include thehighest quality of stainless steel (e.g. 316L).Applications which require very low level detectionof water vapor (e.g. the outlet of molecular sievedriers in cryogenic hydrocarbon processing) shouldbe designed to incorporate passivated, electro-polished stainless steel as a minimum. The use ofelectro-polished tubing reduces the surface areaand roughness of the sample system.

AMETEK generally recommends a samplesystem design which follows a keep it simpleapproach. Any unnecessary tubing lengths, dead

legs (including unnecessary gauges) should beavoided. Sample line lengths, certainly required andnecessary when using extractive analyzer designs,should be kept to a minimum (and heated).

5.3 Sample System TemperatureThe surface coverage (i.e., the total amount of

water bound to the surfaces) of a sample system isa function of temperature. Higher temperaturesresult in lower surface coverage. Insulating andheating sample lines and sample systemcomponents upstream of any sensor or sample cellis highly recommended (AMETEK generallyrecommends a temperature of 60C or higher). Toillustrate this point, a moisture breakthroughcalculation program, developed by Air Products andChemicals Inc., predicts a 100% difference in thebreakthrough time for a moisture event whencomparing a simulated sample line at 0C versusthe same parameters for 60C [7].

While a warm sample system is recommended,it is more important to maintain a stable temperaturein the sample system, from the process connectionto the analyzer sample cell (or analyzer enclosure).Process piping takes up and releases water withambient temperature changes. A sample line with avarying temperature will exhibit the same behavior.If maintaining a sample temperature of 60C is notpossible, then care should be taken to ensure thetemperature is stable, at the very minimum.

5.4 Preventing Contaminat ionNatural gas can be categorized as a dirtysample. The sample gas generally containscomponents which have a likelihood of condensing,and most sensor-based and spectroscopicanalytical technologies presented in this paper aresimply not designed to operate in the presence ofliquids. To ensure the sample is kept in the gasphase, temperature control, as mentioned above,will prevent the natural gas (and any impuritiespresent in the gas) from condensing in the sample.

Further sample conditioning, using membranefilters are a must, as the filters prevent liquid slugsfrom entering the sample system and analyzerduring process upsets. Although liquids havedifferent effects on the various analytical

technologies, any significant quantities of liquidswhich break through and reach an analyzer cell orsensor will result in either damage to the instrument,or require disassembly for maintenance, meaningthe analyzer will be off line until the system iscleaned or repaired.

Coalescing filters may be required to removeadditional impurities such as particulate matter,which may be present in pipeline operations, or dustwhich may be seen at the outlet of a molecularsieve dryer in natural gas processing.


8/8


5.5. Pressure Reduction and Joule-ThomsonCooling

Most analytical technologies designed to monitor

the water vapor concentration in natural gassamples require pressure reduction and control.This is of particular importance for laser-basedsystems, which operate at near atmosphericpressures.

Sample pressure should be kept as low aspossible while maintaining the correct flow. Samplepressure should be reduced as close as possible tothe source of sample extraction. Volumetric turns(i.e. higher sample flow rates) are desired to keepsweeping through the sample line. Surface effectscannot be eliminated, but by keeping volumetric flowhigh, their magnitude can be reduced.

Figure 9 provides an example of a typicalsample system designed for a natural gas service.

Figure 9: Properly Designed Sample System forNatural Gas Moisture Analyzer

And finally, sample system design shouldincorporate design to reduce the effect of Joule

Thomson cooling in natural gas samples.Regulators surrounded by a soccer ball size ball ofice have often been observed in the field. This ball,although dramatic, is not the problem but is a goodindicator. Water molecules stick to cold surfacesbetter than to hot surfaces. Heated, insulatedpressure reduction stages (single or mutli-stage,dpending upon pressure drop requirements), oralternatively using membrane pressure regulatingprobes are required when large pressure drops arerequired to operate water vapor analyzersdownstream of the sample system.

6 SUMMARY

The measurement of water vapor in the naturalgas processing and transmission industry is animportant quality control parameter to ensurenatural gas meets strict pipeline specifications toreduce hazards and maintain pipeline integrity.Among the many analytical technologies availableto determine the water vapour concentration, near-infrared TDLAS devices have found wideacceptance in the industry. Understanding thetechnology, its practical limitations and capabilityhas allowed for the design of sophisticated, reliableand accurate instruments capable of providing fastresponding single, and multi-component

measurement capability.

7. ACKNOWLDEGEMENTS

The author wishes to acknowledge Dr.SamLangridge, TDLAS Product Manager, AMETEKProces Instruments for his invaluable contributionand assistance in writing of his paper.

8. REFERENCES

1. Carroll, J.J. : Natural Gas Hydrates: A Guidefor Engineers Gulf Professional Publishing(2003)

2. Hauer, R., Potter D. Instruments for MeasuringWater and Hydrocarbon Dew Points in NaturalGas. Canadian School of HydrocarbonMeasurement (2004)

3. Potter D. Analytical Devices for theMeasurement of Water Vapor and HydrocarbonDew Point in Natural Gas, Proceedings of the3

rdNatural Gas Sampling Technology

Conference NGSTECH (2011)

4. Amerov, A., Fiore, R., Fuller, M. An InternalVerification Approach for On-Line MoistureAnalysis in Hydrocarbon Gas Streams UsingLaser Spectroscopy. Proceedings of the ISA53

rdAnnual Analysis Division Symposium

(2008)

5. Rothman, L et al, The HITRAN 2004 MolecularSpectroscopy Database Ed. 2004, Journal ofQuantitative Spectroscopy & RadioactiveTransfer 96, 139-201 (2005)

6. Langridge, S. On-Line Measurement ofOxygen: Review and New Developments.Proceedings of the ISA 55

thAnnual Analysis

Division Symposium (2010)

7. Zatko, D; Estimate of Time for Breakthrough ofMoisture in SS EP Tubing version 2.5,copyright Air Products Inc. (1997)

8. Bear, R.; and Blakemore C.; "Reducing the

Detection Limits for a Process MoistureAnalyzer, Proceedings of the 46

thAnnual ISA

Analysis Division Symposium, Volume 34(2001)

9. Amerov, A., Fiore, R., Langridge, S.: ProcessGas Analyzer for the Measurement of Waterand Carbon Dioxide Concentrations,Proceedings of the ISA 54

thAnnual Analysis

Division Symposium (2009)

Documents

GAS 2011 Analytical Devices for the Measure of Water Vapor in the Natural Gas Process and Transmission Industry