The Solution for AccurateGas Flow Measurement PGI International PGI International Excellence Through Innovation Excellence Through Innovation ISO 9001:2000 Certified Quality System Form: TSYNC-NVL Rev. 38.18.08 INTRODUCTION Thermowell Technical Information Reducing Thermowell Conduction Errors in Gas Pipeline Temperature Measurement2 Temperature Error - Can a few degrees really make a difference?6 ThermoSync High Velocity/Pressure Flow Tests, CEESI Iowa7 ThermoSync Test Results:Temp Error vs. Flowrate and Prove Response Time10 Additional Test Results: ThermoSync vs. Conventional Thermowells and PerformanceEvaluations can be obtained by visiting our website at :www.pgiint.com or callCustomer Service at 1-800-231-0233 or 1-713-744-9850 (USA) TECHNICAL AND ORDERING INFORMATION ThermoSync Thermowell Dimensions11 ThermoSync Thermowell Only Ordering Information12 ATP-1000 and ATP-1001 Probe/Sensor13 ATP-1000-R6 and ATP-1001-R6 Probe/Sensor14 ATP High Accuracy RTD Probe/Sensor Only Ordering Information 16 ATP High Accuracy RTD Probe/Sensor Options (photos)17 ATA-Series Ordering Information18 ATA-Series Options (photos)19 USERS LIST Domestic End Users and OEM Users List 20 International End Users and OEM Users List21 INDEX PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 1 Delrin and Teflon are registered trademarks of the E.I. DuPont de Nemours Company. Abstract. It iswell known thatconventionalthermowells will thermallycouple withthe vessel in which they aremountedresultinginanerrorinthetemperaturemeasurement.Conductionerror,commonlycalled immersion error, is present whenever a temperature gradient exist between the vessel or pipe the thermowell is installed into, and the substance being measured. Sources of conduction error in gas temperature measurement and methods of reducing it, specifically the use of finned thermowells, are discussed. In the gas pipeline industry, gas temperature, along withpressureandflow,isusedtocalculate volumetricflow.Anyerrorinthetemperature measurementresultsinanerrorintheflow calculation. This error directly shifts the bottom line, resultinginaccountingfortoomuchgasornot enough.Thisunaccountederrorinthegasvolume measurementcanhavesignificantcostassociated withit.Anytimethereisadifferencebetweenthe pipetemperatureandtheflowinggastemperature therewillbeaconductionerror.Eventhemost accuratetemperaturemeasurementequipmentwill haveerrors.Understandingthesourcesof conduction error and how to minimize it will increase the accuracy of gas temperature measurement. Forthesensortoapproachthetemperatureofthe flowinggastheremustbeathermalcouplingor heattransferbetweenthetwo.Foranetheat transfer from hot to cold to take place there must be atemperaturedifferencebetweenthesensorand thegas.Whenthereisnoheattransfer,the measurementsystemisinthermalequilibriumand thesensorisasclosetothegastemperatureasit willget.Inreality,thesensortemperaturewill neverbethesameasthegastemperaturedue mainly to conduction errors. Reducing Thermowell Conduction Measurement Errors in Gas Pipeline Temperature Measurement Kevin J. Cessac, PGI International INTRODUCTION To accurately measure the flowing gas temperature, the sensor must be in thermal contact with the gas, but not disturb the gas temperature. Figure 1 shows acut-awayviewofatypicalthermowellinstallation inagaspipeline.Athreadedfitting,weldedtothe topofthepipe,providesthemountforthe thermowell.Athermowellfittingconnectsthe thermowelltoaprotectionheadortemperature transmittertoallowfieldwiringtothesensor.The thermowellfittingshouldincludeaspringtokeep thesensorprotectiontubepressedfirmlyintothe thermowellprovidingagoodthermalcontact betweenthesensortipandthebottomofthe thermowell.Thermallyconductivegreaseatthe sensortipwillalsoimproveperformanceasanyair gapbetweenthesensortipandthethermowellwill addtothemeasurementerrorandreducethe response time.THE MEASUREMENT PROCESS FIGURE 1. Typical sensor installation including thermowell and protection head for gas temperature measurement in a pipeline. PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 2 Reducing Thermowell Conduction Measurement Errors in Gas Pipeline Temperature Measurement SOURCES OF CONDUCTION ERROR Heattransfercanbesubdividedintothree categories,conduction,convection,andradiation.Conduction takes place when there is a temperature gradientinasolid,liquid,orgas.Themoleculesin the hot area increase the violence of their vibrations astheyheatup.Then,astheycollidewiththeir slowermovingneighbors,someoftheirenergyof motionisshared,andtheyinturnpassitonfrom one molecule to the next.The equation for thermal conductivity in one dimension is [1]: where qcond is the quantity of heat flow per unit time, Aisthecrosssectionalarea,T2-T1isthe temperature difference, and L is the length.When a metalrodisheatedononeend,heatwillbe conducted through the metal to heat the other end. Convection is the transfer of heat from one place to theotherbytheactualmotionorflowofthe material.The equation for convection heat transfer, also called Newtons Law of cooling is [2]: (2)( ) = T T hA q s convwhereqconvistheheatpowerbeingtransferred throughthe surface, Ais thesurface area, Tsisthe surfacetemperature,Tisthetemperatureofthe fluiddistantfromthesurface,andhisthe convectionheattransfercoefficient(25to250for forcedconvectiongases[3]).Anexampleof convectionheattransferwouldbeafanblowingon ahotobject.Theheatfromthehotobjectis transferredtothecoolerflowingair.Ingas temperaturemeasurement,theheatfromthe thermowell sensing-section and sensor is transferred to the flowing gas mainly by convection. Heatcanalsobetransferredbyradiation,whichis thecontinuousemissionofenergyfromthesurface ofallobjects.Theequationforradiationheat transfer is [2]: whereqradistheradiatedpower,Aisthesurface area,istheStefan-Boltzmannconstant(5.67x10-8Watt/m-2xK-4),TSisthesurface temperatureoftheobject,TSURisthesurface temperatureoftheambientsurroundings.Atthe relativelylowtemperaturesseeninnaturalgas pipelines,typicallylessthan70C,therateof radiationheattransferissmallandisatlong wavelengthsproducinglittleinfluenceonthe measurement accuracy.Radiation heat transfer can beminimizebymanufacturingthethermowellofa lowemissivitymaterial,suchaspolishedstainless steel with emissivity = 0.17 [4]. ( )4 4SUR Srad T T A q = (3) HEAT TRANSFER LT TkA qcond1 2 =(1) FIGURE2.Cutawaypipesectionshowing(a)sensorin flowinggaswithoutconductionpath,(b)conductionpath throughthesensorwires,and(c)largeconductionpath through thermowell, protection tube, and sensor wires. Conductionerrorsarepresentwheneverthereisa differenceintemperaturebetweentheflowinggas andthepipelinewalls.Thepipelinerunning underground is kept at a fairly constant temperature but at the metering station, a section of the pipeline isbroughtabovegroundtoallowaccessforthe measurementequipment.Themeteringsectionof pipeandtheequipmentinstalledonthepipe,are heatedandcooledbythechangingambient environment.Asthetemperaturedifference increases,theconductionerrorforanyinstallation also increases. To understand the sources of conduction errors, first consideratemperaturesensorinsertedintothegas stream as shown in Figure 2a. PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 3 Reducing Thermowell Conduction Measurement Errors in Gas Pipeline Temperature Measurement Thesensorisonlyinfluencedbyconvectionanda smallinfluencefromradiationheattransferandwill eventuallybeequaltoorverynearthegas temperature.Thereisnoconductionerrorbecause there is no conduction path to the sensor.In Figure 2b,sensorwiresareadded,whichcreatesa conductionpathbetweenthewarmpipeandthe sensor.Heatisconductedfromthepipewalls throughthewirestothesensor,andactuallyshifts thesensortemperature,resultinginaconduction error. As long as there is a temperature difference between the pipe and the gas, the sensor will never equal the gastemperatureduetothethermalenergy conductedthroughthesensorwires.InFigure2c, thethermowell,sensorprotectiontube,and protectionheadareaddedtothesensor.This significantly increases the cross sectional area of the conduction path to the sensor.Referring to Equation 1,theincreasedcrosssectionalareaincreasesthe conductionheattransfertothesensor,increasing the conduction error. If the pipe was cooler than the gas,thesameerrorexistsforthesamedifferential temperaturebutintheoppositedirection.The followingrelationmaybeusedtodeterminethe extent of the conduction error [5]: Themostcommonmethodforminimizing conduction errors is to increase the insertion length.Typicallytheinsertionlengthofthethermowell should equal a minimum of 10 times the diameter of thethermowell[6].Increasingtheinsertionlength will reduce conduction errors as shown in Equation 4 butinsmalldiameterpipesthismaybeimpossible.Reducingthethermowelldiameter,toreducethe conductionpath,isnotpracticalinhighpressure, highflowvelocitypipelinesbecausethestrengthof the well will be compromised.Flow rates can be as highas30meterspersecondwithpressuresupto 7,000 kPa. Reducing the thermal conductivity of the thermowell material (k in equation 4) will reduce the conducted thermalenergybetweenthepipewallsandthe sensor,resultinginareducedinfluencetothe measurement.Thedrawbackisthatitwillalso reducethethermalconductionthroughthesensing section of the thermowell to the sensor, reducing the influenceofthecoefficientofheattransfer(hin equation4) andreducingthesensorresponsetime. Thethermowellmustbemanufacturedofmaterials compatiblewiththeapplication.Tominimize radiationheating,thematerialshouldhavealow emissivity surface. Typically thermowells used in gas pipelinetemperaturemeasurementaremadeof polished stainless steel, a tradeoff between accuracy and strength. Furtherexaminationofequation4indicatesthatif the surface area of the thermowell at the sensor was increased(pinequation4),withoutincreasingthe conductioncrosssectionalarea(ainequation4)or thethermalconductivity(kinequation4), conductionerrorswillbereduced.Thiscanbe accomplishedbyaddinganarrayoffinsatthe sensing section of the thermowell.Figure 3a shows aconventionalthermowellforcomparisontoa thermowellwiththeaddedfinstoincreasethe surfaceareaofthesensingsection,Figure3b.Adding the 13 fins increases the surface area in the sensing section over 7.5 times that of a conventional thermowell. (4) =) (mL Cosh T TT Tsg wi sgwhereTsgisthestatictemperatureofthegas,Tiis the indicated sensor temperature, Tw is the pipe wall temperature, L is the immersion length, m = (hp/ka)1/2 , h is the convection coefficient of heat transfer, p is the surface area of the thermowell (at the sensor), k is the thermal conductivity of the thermowell, and aistheconductioncross-sectionalareaofthe thermowell.The larger mL becomes, the closer the indicatedtemperature,Ti,approachesthetruegas temperature,Tsd.AnymeansofincreasingthemL productwillresultinareducedconductionerror.Thesteadystatetemperatureofthesensorisa result of a balance between convection heat transfer from the gas to the thermowell, and conduction and radiationheattransferbetweenthesensorandits surroundings. REDUCING CONDUCTION ERROR PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 4 Reducing Thermowell Conduction Measurement Errors in Gas Pipeline Temperature Measurement FIGURE 3.(a) Conventional thermowell(b) Finned thermowell The increased surface area at the sensing section of thefinnedthermowellsignificantlyincreasesthe convectionheattransferbetweenthegasandthe sensor, shifting the sensor temperature closer to the actualgastemperature,reducingtheconduction error.Equation2showsthatanincreaseinthe surfacearea,A,willincreasetheheatpower transferred, qconv. FINNED THERMOWELL PERFORMANCE FIGURE4.Flowcomparisonofaconventionalthermowell, top curve, and a finned thermowell, bottom curve. [7] Actualflowtestingclearlyshowstheperformance improvementstheincreasedsurfaceareaprovides.Aconventionalthermowellandafinnedthermowell wererunina7.5cminsidediameterschedule80 pipe.Thepipewasheatedto38C andheldatthis temperature while various flow rates were observed. Figure4showsthedifferencebetweentheactual gastemperatureandtheindicatedtemperature.Thefinnedthermowell,indicatedbythebottom trace, was very close to the actual gas temperature.Theconventionalthermowellhadover1.6C(3F) errorovertheentiretest.Slowerflowvelocities producedlargererrorsasexpected.Sensor response time was four times faster using the finned thermowell [7]. REFERENCES 1.SearsF.W.,ZemanskyM.W.,YoungH.D.,College Physics Fifth Edition 1980. 2.BentlyR.E..TemperatureandHummidity Measurement, Volume 1, 1998. 3.IncroperaF.P.,DeWittD.P.,editors.Fundamentalsof Heat and Mass Transfer. John Wiley & Sons, Inc. New York 1990. 4.OS550IndustrialInfraredThermometerUsersGuide, Emissivity Table (Omega Engineering). 5.ASMETemperatureMeasurement,(AmericanSociety of Mechanical Engineers), PTC 19.3 1974, reaffirmed 1998, Part 3. 6.Liptak B.G., editor-in-chief, Process Measurement and Analysis, chapter 4. 7.TAN-34CO-L4ThermowellPerformanceEvaluation, ThermoSyncTemperatureMeasurementSystem,PGI International, October 2000. PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 5 Can a few degrees really make a difference? Using the flow lab test results, two flow rates were used to compare the conventional stainless steel thermowell to the ThermoSync temperature measurement system.At 60 CFM the temperature error for the conventional system is 7.5F and at 300 CFM 3.8F.For the ThermoSync system, 0.2F at 60 CFM and 0.0F at 300 CFM.These are conservative errors taken with only a 30F difference between the pipe temperature and actual gas temperature.In extreme ambient conditions, where the difference is larger, expect larger temperature errors.The two flow rates were converted to SCFH using 200 PSI at 70F. 60 ACFM = 48 MCF/Hour 300 ACFM = 240 MCF/Hour Using AGA3 equations, the following parameters were used to determine the effect of the temperature error on a calculated flow volume for an orifice plate flow meter. Flange TapsUpstream Pressure Base14PSIA Temp Base 60F Barometric Pressure14.7 Static Press.200 PSIG Pipe I.D.2.97" Orifice1.50" Specific Gravity1.00 CO2 Mole Percent0.00 N2 Mole Percent0.00 Flow Rate ThermowellTempMCF/Hour MCF Error/Hour Dollar Error/Hour Dollar Error/24 Hours Dollar Error/31 days 60 CFM Actual70.048.0030$ 0.00$ 0.00$ 0.00 ThermoSync70.247.9910.012$ 0.11$ 2.59$ 80.35 Conventional77.547.5840.419$ 3.77$ 90.50$ 2,805.62 Actual70.0240.0000 $ 0.00$ 0.00$ 0.00 ThermoSync70.0240.0000.000$ 0.00$ 0.00$ 0.00 Conventional73.8238.9311.069$ 9.62$ 230.90$ 7,158.02 300 CFM Dollar error per hour for the two temperature measurement methods (based on $9.00 per MCF) Even under these conservative conditions, a conventional thermowell may be costing you $230.90 per day...or $84,278.50 per year! $9.62 x 24 hours per day = $230.90 $230.90 x 365 days per year = $84,278.50 Temperature Induced Error in Turbine or PD Meters Flow Rate ThermowellTempMCF/Hour MCF Error/Hour Dollar Error/Hour Dollar Error/24 Hours Dollar Er-ror/31 days 60 CFM Actual70.048.1440$ 0.00$ 0.00$ 0.00 ThermoSync70.248.1240.020$ 0.18$ 4.32$ 133.92 Conventional77.547.4040.740$ 6.66$ 159.84$ 4,955.04 Actual70.0239.9810 $ 0.00$ 0.00$ 0.00 ThermoSync70.0239.9810.000$ 0.00$ 0.00$ 0.00 Conventional73.8238.0951.886$ 16.97$ 407.38$ 12,628.66 300 CFM Todeterminethemeasurementerrorinultrasonic,turbineandPDmeters,theSCFMiscalculatedusingthe actual gas temperature.This rate is compared to the SCFM calculated using the measured temperatures from the ThermoSync system and the conventional system. Even under these conservative conditions, a conventional thermowell may be costing you $407.28 per day...or $148,657.20 per year! $16.97 x 24 hours per day = $407.28 $407.28 x 365 days per year = $148,657.20 PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 6 ThermoSync Flow Stability Test Report High VelocityHigh Pressure Flow Test, CEESI Iowa TAN-34CO-L2, TAN-34CO-L4, TAN-34CO-L8, TAN-34CO-L12 January 15, 2001 Designequationsforthestrengthofathermowell,naturalfrequency,andthewakeorStrouhalfrequencyof thermowellshavebeenavailableforsometime,andarewell-documented(ASMEPTC19.3Temperature Measurement).Theseequationshowever,aredevelopedaroundaconventionalthermowellprofileanddonot apply to the finned ThermoSync design.To determine the application limitations for the ThermoSync thermowells, testingatthemaximumoperatingconditionsmustbedone.Itwasdeterminedthatthemaximumpressure normally seen in a natural gas pipeline is 1000 PSI.Similarly, the maximum flow velocity was determined to be 100 feet per second.Working with CEESI Iowa natural gas flow measurement facility in Garner Iowa, we were able to define the test requirements. BACKGROUND TheColoradoEngineeringExperiment Station,Inc.(CEESI)isanindependent commercial calibration and flow research facility.CEESIalsoparticipatesinflow measurementstandardscommittees organizedbytheASME,AGA,API,and GRI.Many years of experience in meter provi nghavegai nedCEESIthe internationalreputationasaleaderin flowmeasurement.TheCEESIIowa highflowtestfacilityhastheunique abilitytoflownaturalgasupto20,000 ACFMat1050PSIinacontrolledtest environment.CEESI Tosimplifythetest,aspoolwas designedtoallowsixthermowellstobe installed and tested together.The spool is a 10 schedule 40 pipe with ANSI 600 lb. flanges as shown in figure 1. TEST SET-UP FIGURE 1. Test spool design. FIGURE2.Testspooland thermowells installed in CEESI flow test pipe. PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 7 ThermoSync Flow Stability Test Report ThethermowellsinlocationsC,D,E,andFwerefittedwithEntranminiatureEGAaccelerometerstomeasure vibration.Each accelerometer shared the same amplifier/power supply with a gain of 41.6 making the output of each sensor 50mv/g(1.2mv/g x 41.6 = 50mv/g). Accelerometer Specifications ModelEGA-250-/R Range+/-250g Limit+/-1250g Temp. Range-40 to 250F Non-Linearity+/-1% OutputApprox. 1.2mv/g TheconventionalthermowellinlocationA(withaconventionalRTDprobe)andtheThermoSyncthermowellin location B (with the ATP-1000) were used for a temperature accuracy comparison.Both probes were connected to a Rosemount 3244MV transmitter.The transmitter was setup to provide an output proportional to the difference of the two temperatures.If both sensors are at the same temperature, no thermowell error, the temperature output will be 12ma (1/2 scale).Any temperature error will shift the transmitters output.While running the test, the pipe temperaturewasafairlyconstant60Fandthegastemperature69.23F(measuredbyCEESI).Witha9F difference between pipe wall and gas, the measured temperature difference was between 0.1F and 0.2F at the 50ft/sec. flow rate. Flow velocity = 100 feet per second Static Pressure.............................................................................................. 1080 PSI TAN-34CO-L2 ...................................................................................................... 8.8g TAN-34CO-L4 ...................................................................................................... 9.2g TAN-34CO-L8 .................................................................................................... 10.4g TAN-34C0-L12................................................................................................... 11.2g All of the thermowells produced only low level background vibration at 100 ft/sec. Flow Velocity = 112 feet per second Static Pressure.............................................................................................. 1080 PSI TAN-34CO-L2 ....................................................................................................... 22g TAN-34CO-L4 ....................................................................................................... 20g TAN-34CO-L8 ..................................................................................................... 112g TAN-34C0-L12................................................................................................. > 250g Increasing the flow velocity from 100 to 112 feet per second resulted in L12 breaking into oscillation at 540 Hz.Most of the vibration in the other wells was produced by L12 interference.This is clearly shown by the datatakenwithL12removed.WhenthewakeorStrouhalfrequencyofawellapproachesthenatural frequency,thethermowellbreaksintooscillations.Anaudibletoneatthevibrationfrequencyclearly indicated when this happened.The two waveforms below show the L12 thermowell at 100 ft/sec and 112 ft/sec. TEST RESULTS PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 8 ThermoSync Flow Stability Test Report Flow Velocity = 112 feet per second, L12 Removed Static Pressure.............................................................................................. 1080 PSI TAN-34CO-L2..........................................................................................................8g TAN-34CO-L4..................................................................................................... 11.2g TAN-34CO-L8..................................................................................................... 14.4g L12 was removed from the spool to allow the other probes to be measured at higher flow velocities. Removing L12 reduced the vibration levels measured on the other wells. Flow Velocity = 135 feet per second, L12 Removed Static Pressure.............................................................................................. 1080 PSI TAN-34CO-L2..................................................................................................... 10.4g TAN-34CO-L4..................................................................................................... 17.6g TAN-34CO-L8..................................................................................................... 25.6g Increasing the flow velocity slightly increased measured vibration levels. Flow Velocity = 150 feet per second, L12 Removed Static Pressure.............................................................................................. 1080 PSI TAN-34CO-L2..................................................................................................... 15.2g TAN-34CO-L4..................................................................................................... 31.2g TAN-34CO-L8........................................................................................................ 28g At 150 feet per second, there is no excessive vibration. TheL12thermowellisunacceptableforuseabove100feetpersecond.TheL2,L4andL8thermowells performed without excessive vibration up to the 150 ft/sec and should provide long-term performance in flow velocities up to 100 feet per second. TheL12shouldbetestedagainwithouttheotherprobesinstalledinthespooltodetermineifthereis interference from the other thermowells.Notice in Figure 1 that the other probes may interfere with the L12 flow profile. The measured temperature error between the ThermoSync L4 system and a conventional thermowell / probe (3 insertion length) at 50 feet per second was approximately 0.2F.Calculating the volume using 69.23F and 69.03F, there could be a 5.7 MCF per hour error in the calculated flow volume due to pipe wall thermal coupling in the thermowell.At a gas price of $9.00 per MCF, thats a $51 per hour, $1,231 per day or$449,388peryearunaccountableerror. Using the ThermoSync system can significantly reduce pipe wall coupling errors. SUMMARY PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 9 ThermoSync Test Results Conventional Thermowellwith Stainless Steel ProbeTAN-34C0-L4 Thermowellwith ATP-1000 ProbeTEMPERATURE ERRORvs. FLOW RATE 3 schedule 80 pipe at 100F with a nominal supply of air temp of 70F at 0 PSIG The zero error readings at zeroCFMshowthetest pipeandtemperature sensorshavestabilizedat 100F.Astheflowrate increasesthetemperature oftheairattheair t emper at ur esensor quicklydropstothe nominalairtempof70F showingalargeerror.With continued increase in flowrate,theerrordrops as the cooling effect of the flowing gas overcomes the heatinginfluenceofthe pipe.Theerrorcontinues droppingwithincreasing flowrate,butlevelsout substantiallybeyond200 CFM. Attimezerowiththe temperaturesensorsand thetestpipestabilizedat 100Fthereisnoerror.Theflowratewas suddenlychangedto100 CFMcausingasudden maximum error.The flow rateisheldat100CFM throughthedurationof thetest.Theresponse timefortheThermoSync thermowellisabout120 seconds.Forthe conventionalthermowell, responsetimeisabout 480seconds.Theprobe typehadnoeffecton responsetimeofeither thermowell.PROBE RESPONSE TIME 3 schedule 80 pipe at 100F with a nominal supply of air temp of 70F at 0 PSIG At time zero, test pipe and temperature sensors have stabilized at 100F.At time zero, flow is switched from 0 to 100 CFM.Response Time 3-inch Schedule 80 Pipe at 100F with a 70F nominal supply air at 0 PSIG100CFM Error, Degrees Fahrenheit Flow, Cubic Feet / Minute Seconds Error, Degrees Fahrenheit PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 10 ThermoSync Dimensions Process Connection PART NUMBERABCDEFGHIJKLMStem TAN-12C0-L21/2 FNPT1.251/2 MNPT1.690.633.4952.221.20.260.37.6453.88.90Stepped TAN-12C0-L41/2 FNPT1.251/2 MNPT1.690.633.4952.961.20.260.37.6454.75.90Stepped TAN-12C0-L81/2 FNPT1.251/2 MNPT1.690.633.4954.591.20.260.37.6456.37.90Stepped TAN-12C0-L121/2 FNPT1.251/2 MNPT1.690.633N/A6.661.20.260.37.6458.45.90Straight TAN-12C0-L241/2 FNPT1.251/2 MNPT1.690.633N/A9.891.20.260.37.64511.67.90Straight TAN-34C0-L21/2 FNPT1.253/4 MNPT1.690.808.4952.221.20.260.37.853.82.90Stepped TAN-34C0-L41/2 FNPT1.253/4 MNPT1.690.808.4952.961.20.260.37.854.56.90Stepped TAN-34C0-L81/2 FNPT1.253/4 MNPT1.690.808.4954.591.20.260.37.856.20.90Straight TAN-34C0-L121/2 FNPT1.253/4 MNPT1.690.808N/A6.661.20.260.37.858.26.90Straight TAN-34C0-L241/2 FNPT1.253/4 MNPT1.690.808N/A9.891.20.260.37.8511.49.90Straight TAN-10C0-L41/2 FNPT1.3751 MNPT1.690.808.4952.961.20.260.37.854.75.90Stepped TAN-10C0-L81/2 FNPT1.3751 MNPT1.690.808.4954.591.20.260.37.856.37.90Stepped TAN-10C0-L121/2 FNPT1.3751 MNPT1.690.808N/A6.661.20.260.37.858.45.90Straight TAN-10C0-L241/2 FNPT1.3751 MNPT1.690.808N/A9.891.20.260.37.8511.67.90Straight All Thermowells: Material:Press/Temp:Flow: 316L SS 4900 PSI Max @ 330 F 100 FPS (L2, L4, L8, L12) or 50 FPS(L24) max in 1000 PSI Natural Gas NOTE:Use a thermal coupling paste or fluid to couple the probe tothewellONLYinthelower.5inchesofthewell.DONOTfill thewellwiththermalcouplingfluid.Springloadtheprobeto contact the bottom of the well. CRN:OH7489.5123467890YTN PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 11 Thermowell Only Ordering Information PROCESS CONNECTION SIZEMATERIALFITS PIPELINE SIZE OPTIONS TAN-XXCO--XXXXXXXXXXPROCESS CONNECTION SIZE CODEDESCRIPTION 121/2 MNPT Process Connection 343/4 MNPT Process Connection 101 MNPT Process Connection(not available for the 2 Pipeline Size) MATERIAL CODEDESCRIPTION CO316L Stainless Steel FITS PIPELINE SIZE CODEDESCRIPTION L22 Pipeline L43 - 4 Pipeline L86 - 8 Pipeline L2416 - 24 Pipeline L1210 - 12 Pipeline OPTIONS CODEDESCRIPTION PCap & Chain Assembly(Can be ordered separately as Spare Part # ATA-1002) VVent/Bleed Hole in Thermowell (1/8) W1316 SS Identification Tag Process Connection 300 SSClip Ring 300 SSChain OPTIONS PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 12 Aluminum Cap P OptionViton O-Ring 1/8 Vent Hole V Option ATP-1000 & ATP-1001 Probe/Sensor The ATP-1000/1001 RTD probes reduce thermal coupling errors from thermowells by using a unique two-piece protection tube.The body of the protection tube is made of insulating PVC plastic, thermally insulating the RTD sensorfromthethermowellwalls. Thishelpsblock thepipethermaltransferdowntheprotection tubetothe sensor.The RTD sensor is mounted in a thermal conductive body for best accuracy and response. For maximum accuracy, the ATP-1000 and ATP-1001 probes should be used with a TAN ThermoSync thermowell. SPECIFICATIONS 4-wire platinum wire-wound RTD Element 100 Ohms at 0 C (IEC 751) PT 100 typical calibration coefficients (ITS-90) A (C)3.90830 x 10-3 B (C)-5.77500 x 10-7 C (C)-4.18301 x 10-12 Alpha (C)3.850 x 10-3 Delta (C)1.49990 Beta (C)0.10863 Band 5 sensor, interchangeability accuracy 0.025C at 0C, 0.07C at 50C1 Calibration and accuracy certification service available -40C (-40F) to +60C (+140F) temperature range Sensor has high thermal conductivity hard-anodized aluminum tip for maximum thermal transfer between the thermowell and sensor PVC protection tube thermally insulates the thermowell neck from the sensor 26 AWG Teflon leads Probe length user-cut to fit application Probe body is laser marked with serial number and ratings for calibration traceability SENSOR LEAD CONNECTIONS The4-wireprobecanbeusedwith4-wire,3-wire and2-wireRTDtransmitters.For3-wiresystems, use two red leads and one white lead.Insulate the unused white lead (do not connect the unused white lead).For 2-wire systems, connect both red leads to oneterminalandbothwhiteleadstotheother terminal. PROTECTION TUBE CUTTINGPROCEDURE 1.Determinetheprobelengthrequiredforyour installation.Besuretoleaveroomforthe spring-loaded fitting. 2.Usingasmalltubingcutter,scoretheplastic tubingabouthalf-waythrough.Toprevent damagetothesensorleads,nevercutthrough the plastic protection tube with the tubing cutter.Minimum probe length is 2 inches. 3.Removethetubingcutterandapplybending pressure at the score mark.The protection tube shouldbreakatthescoremark.Ifexcessive pressureisrequired,usethetubingcutterto make the score deeper. PATENTED PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 13 ATP-1000-R6 & ATP-1001-R6 Dual Probe/Sensor The ATP-1000/1001-R6 RTD Dual RTD probes reduce thermal coupling errors from thermowells by using a unique two-piece protection tube.The body of the protection tube is made of insulating PVC plastic, thermally insulating the RTD sensors from the thermowell walls.This helps block the pipe thermal transfer down the protection tube to the sensors.The RTD sensors are mounted in a thermal conductive body for best accuracy and response. Formaximumaccuracy,theATP-1000-R6andATP-1001-R6DualRTDprobesshouldbeusedwithaTAN ThermoSync thermowell. SPECIFICATIONS Wire-wound RTD Elements 100 Ohms at 0 C (IEC 751) PT 100 typical calibration coefficients (ITS-90) A (C)3.90830 x 10-3 B (C)-5.77500 x 10-7 C (C)-4.18301 x 10-12 Alpha (C)3.850 x 10-3 Delta (C)1.49990 Beta (C)0.10863 Band 5 sensor, interchangeability accuracy 0.025C at 0C, 0.07C at 50C1 Calibration and accuracy certification service available -40C (-40F) to +60C (+140F) temperature range Sensor has high thermal conductivity hard-anodized aluminum tip for maximum thermal transfer between the thermowell and sensor PVC protection tube thermally insulates the thermowell neck from the sensor 26 AWG Teflon leads Probe length user-cut to fit application Probe body is laser marked with serial number and ratings for calibration traceability SENSOR LEAD CONNECTIONS The Dual RTD probe can be used with 3-wire and 2-wire RTD transmitters.For 3-wire systems, use the two red leads and one white lead for the first sensor, and the two orange leads andoneyellowleadfor thesecondsensor.For 2-wiresystems,connect bothredleadstoone terminalandthewhite lead to the other terminal forthefirstsensor,and connectbothorange leads to one terminal and theyellowleadtothe other terminal for the second sensor. PROTECTION TUBE CUTTINGPROCEDURE 1.Determinetheprobelengthrequiredforyour installation.Besuretoleaveroomforthe spring-loaded fitting. 2.Usingasmalltubingcutter,scoretheplastic tubingabouthalf-waythrough.Toprevent damagetothesensorleads,nevercutthrough the plastic protection tube with the tubing cutter.Minimum probe length is 2 inches. 3.Removethetubingcutterandapplybending pressure at the score mark.The protection tube shouldbreakatthescoremark.Ifexcessive pressureisrequired,usethetubingcutterto make the score deeper. PATENTED PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 14 ATP-1000 RTP Probe/Sensor CalibrationSerial # 1634 TEST EQUIPMENT: ModelDescriptionSerial NumberCalibration Date Calibration DueHart 15216-1/2 Digit ThermometerA13579 5/16/2001 Hart 7102-5 to 125 C Micro-BathA13424 2/12/2001 Agilent 34401A6-1/2 Digit Multimeter US3613095711/19/200311/19/2004 TEST MEASUREMENTS: Temperature FTemperature C Resistance 32.9570.532100.185 75.04323.913109.292 89.98432.213112.516 119.95848.866118.949 Temperature 22C / 72F Date 8/13/2004 CALLENDARVAN DUSEN COEFFICIENTS: CoefficientIPRT Typical ** Calibrations Coefficients R @C100.00099.969 Alpha0.003850.00385698 Delta1.49991.47929 Beta0.10863* A3.908300 E-033.91404 E-3 B-5.77500 E-07-5.70561 E-7 C-4.18301 E-12-4.18984 E-12 Calibration accuracy:+/- 0.025C or +/- 0.05F *The beta coefficient is only used for temperatures below C.The value for the beta coefficient is not solved because there are no samples below 0C.Use the Typical beta coefficient. **Coefficients IAW ITS90. PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 15 ATP High Accuracy RTD Probe/Sensor Only Ordering Information PROBE/SENSOR LENGTHOPTIONS ATP-XXXX-XXXXXXX PROBE/SENSOR LENGTH CODEDESCRIPTION 100012.60 Probe/Sensor Length 100118.60 Probe/Sensor Length All probe bodies are laser marked with a serial number and ratings of traceability. User cut to length with tubing cutter. OPTIONS CODEDESCRIPTION AArmored Cable (Valid only with LXX Option) CRTD Calibration (4-Point or 6-Point, depending on RTD Selected) L1212 foot Shielded Cable Leads with Weather-Tite Connector Seal L2424 foot Shielded Cable Leads with Weather-Tite Connector Seal L5050 foot Shielded Cable Leads with Weather-Tite Connector Seal NXXNon-Standard RTD Length (Valid only with LXX Option; 1 Increments; Minimum Length 2) R66-Wire Dual RTD Probe SSpring Loaded (Includes Spring, Locking Collar and 1/2 MNPT Fitting) W1316 SS Identification Tag W1C316 SS Tag with Calibration data (Valid only with C option) PATENTED PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 16 High Accuracy RTD Probe/Sensor Options Standard 12 Shielded Cable 316 SSIdentification Tag No CableSpring-Loaded Only ATP-1000-L12SW1 ATP-1000-S PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 17 ATA-Series Ordering Information FITS PIPELINE SIZEOPTIONS ATA-XXXXXXX-XXXXXXX PROCESS CONNECTIONCODEDESCRIPTION 10003/4 MNPT Process Connection 20001/2 MNPT Process Connection PROCESS CONNECTION 30001 MNPT Process Connection (not available for the 2 Pipeline Size) FITS PIPELINE SIZE CODEDESCRIPTION L22 Pipeline L43 - 4 Pipeline L86 - 8 Pipeline L1210 - 12 Pipeline L2416 - 24 Pipeline OPTIONS CODEDESCRIPTION BExplosion-Proof Conduit Box (includes Connector for RTD1/2 NPT Connections) CRTD Calibration (4-Point or 6-Point, depending on RTD Selected) F316 SS RTD Fitting (Delrin Standard) PCap & Chain Assembly(Can be ordered separately as Spare Part # ATA-1002) R44-Wire Spring-Loaded RTD Probe R66-Wire Spring-Loaded Dual RTD Probe U316 SS Union (PVC Standard) VVent/Bleed Hole in Thermowell (1/8) W1316 SS Identification Tag W1C316 SS Tag with Calibration data (Valid only with C option) ATA Base Models Consist of:316 SS ThermoSync Thermowell, 316 SS Close Nipple, PVC Union and Delrin Coupler PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 18 ATA-Series Options ATA-1000L4-R4 ATA-1000L4-R4U ATA-1000L4-BR4U PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 19 Thermosync Thermowell RTD Probe/Sensor Gas Temperature Measurement End Users (USA) ATMOS Energy CorporationMississippi Coastal Oil & GasTexas Coastal Flow MeasurementTexas Devon Energy /Devon Gas ServiceTexas, Oklahoma Blue Water Constructors, Inc.Florida El Paso Natural Gas CompanyTexas, New Mexico, Louisiana Eott EnergyCalifornia Florida Gas TransmissionFlorida, Texas Gemma Power Systems, LLCFlorida Kinder Morgan Inc.Colorado, Wyoming, Texas, New Mexico, Illinois Northern Border Pipeline CompanyMinnesota, Indiana, Illinois Northern Natural Gas CompanyAlabama, Kansas, Minnesota,Louisiana, Nebraska, New Mexico,Florida, Iowa, Texas PanHandleEastern Pipeline Kansas, Texas, Oklahoma Peoples GasFlorida Reliant Energy Field Services, Inc.Louisiana Sierra Pacific PowerNevada Transwestern PipelineNew Mexico Questar Gas Pipeline / TransmissionWyoming, Utah Western Gas ResourcesWyoming XTO EnergyTexas, Oklahoma, Alaska, Colorado, New Mexico, Wyoming, Louisiana, Utah, Kansas OEM Accounts (USA) Automation XNew Mexico Instromet, Inc. Texas J-W Measurement Company Colorado, Texas J.W. Williams-FlintWyoming Red Man Pipe & SupplyOklahoma Thermo ElectronTexas Wedge Measurement & ControlTexas Domestic End Users & OEM Users List PGI International 16101 Vallen Drive Houston, TX 77041 713-466-0056 1-800-231-0233 www.pgiint.com 20 16101 Vallen Drive Houston, Texas 77041 USA 713.466.0056Phone 800.231.0233Toll Free 713.744.9897Fax sales@pgiint.com www.pgiint.com PGI InternationalAugust 2008 PGI International PGI International Excellence Through Innovation Excellence Through Innovation INSTRUMENTATION PRODUCTS Instrument Valves & Manifolds Power and Steam Plant Valves & Manifolds Purge Adapters for the Process Industry ENGINEERED PRODUCTS Gas & Liquid Sampling Systems Natural Gas Sampling System Heated Enclosures Sample Cylinders and Accessories MEASUREMENT ACCURACY PRODUCTS ThermoSyncTM Thermowells & Temperature Probes Direct-MountTM Systems Square Root Error (SRE) & Gauge Line Error (GLE) Indicators ZEUS POWER GENERATION SYSTEMS TECTM ThermoElectric Battery Chargers DB1TM Differential Pressure Battery Chargers ADDITIONAL PGI INTERNATIONAL PRODUCTS & SERVICES Valve Fittings & Wellhead Components Propane and Anhydrous Ammonia Valves Contract Machining