Purdue University Purdue e-Pubs
International Compressor Engineering Conference School of Mechanical Engineering
1976 Determining the Air Content in Small Refrigeration Systems V. Gustafsson
Follow this and additional works at: https://docs.lib.purdue.edu/icec
Gustafsson, V., "Determining the Air Content in Small Refrigeration Systems" (1976). International Compressor Engineering Conference. Paper 232. https://docs.lib.purdue.edu/icec/232
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] for additional information. Complete proceedings may be acquired in print and on CD-ROM directly from the Ray W. Herrick Laboratories at https://engineering.purdue.edu/ Herrick/Events/orderlit.html DETERMINING THE AIR CONTENT IN SMALL REFRIGERATION SYSTEMS
Vibeke Gustafsson
Danfoss, Nordberg, Denmark
INTRODUCTION DEFINITIONS For many reasons the presence of air in In the following, references a refrigeration to refrigera system is unfortunate. tion systems are taken as Not meaning complete only does it reduce the capacity of units in domestic use: the refrigeration system, but more impor tant, it destroys the valve system of the Refrigerators compressor, since it promotes the formati Deepfreezers on of carbon i.e. during the oxidative bre Air conditioning units ak down of the refrigerator oil. In a her etc. metically sealed Freon refrigeration cir cuit using a compressor lubricated with mi Described here neral is the way in which the NC oil there exists - other things being gas content in equal such complete systems is de - a simple correlation between air termined. The term content and NC-gas means N nitrogen compressor life time. o oxygen and Ar Argon, 2 It is therefore 2 and it is short for in the interests of both non condensable gas. the compressor manufacturer and the appli ance manufacturer to minimize the air con tent in a refrigeration system. NC-GASSES IN REFRIGERATION SYSTEMS To do this effectively, it is necessary to have a measuring procedure for determi As an introduction a resume of ning the air content. Gas where the chromatographic NC-gasses in a refrigeration system analysis of refrigerants in connection come from and what damage they produce is given. with life tests on compressors and the ma The description is a little brief but ser terials used in their manufacture have been ves only to indicate the necessity of gas known and used for several years by com analysis. pressor manufacturers, but so far this technique has not been applied to any great In our opinion the most important source extent for rutine control of air of in finis failure is the evacuation of the assembled hed refrigeration systems. We use gas system chromatography before charging with Freon. for such measurements, and In a refrigerant find it system the free gas volume is a suitable method in connection typically amounts with modern refrigeration to 2-3 1, and the air system assembly content in this volume line: It is quick and requires has to be removed. only little With plainly deficient or just deficient mannpower. The equipment ne·cessary is in evacuation expensive the complete system will contain to procure and easy to maintain. air; NC-gas. The technique is generally known and re But this cognized. can also be found in a well evacu ated, finished refrigeration system partly This article describes how the absolute that which is not evacuated from the oil air quantity (N , 0 and Ar) in a 2 2 refrige partly that which is charged together with ration system is deEermined by first de the refrigerant. As is well known oil can termining the vol % of NC-gas (by means of contain air and water which is difficult to gas chromatography) in a small sample of remove (1), but these amounts are so small the gas charge in the system and then, on that they the will hardly cause failure. basis of this and knowledge of other If the refrigeration system data, system is charged with calculating the total air con Freon liquid direct tent. Finally, from supplyers' tanks an error estimation is made -via a charging hoard for the method. - any NC-gas which can enter the system will be very small. On the other hand it is our experience that
480 it is just these charging processes which Gas chromatography is an other way of ma are often critical, because of leaks, line king such determinations. connections etc. This method is a normal and reliable one As a final source of failure we can consi- for whole or part separations of gas mixtu der how well the refrigeration system is res in individual components. The more dif sealed. A system which sooner or later de ferent these components are the easier it velops leaks on the pressure side will is to separate them one from another by loose its charge. If leaks appear on the gas chromatography. suction side the compressor will periodi In the case of NC-gas in Freon it is fairly cally suck in air. Air diffusion due to easy to separate the NC-gasses from Freon, partial pressure differences must also be because the individual gas molecules are horned in mind. very different in size and polarity. The principle of gas chromatography is as Hhen the compressor is not running the air follows: 3 in the system will not cause damage. The A small sample - one or more em - is pla oil and the other materials used are com ced on a column, which has been filled with pletely stable in air at room temperature. a suitably porous material. A carrier gas flows constantly through the column carry When the refrigeration system is put into ing the sample with it. Each sort of mole operation and the compressor begins to run cules in the sample will move through the the presence of air will produce two speci column at a different speed and separation al effects: will occur. The first will influence system capacity. Each component of the sample is lead by the The compressor and condenser are not effi carrier gas into a detector. ciently utilized. The degree of influence The most suitably is a hot Wire detector or will depend very much on how much of it similar, i.e. an electric bridge concisting there is in the system in relation to sy of two platinium wires havj_ng the same e stem dimensions and load. With a rising lectrical resistence, but each placed in content of NC-gas, discharge pressure rises its own chamber (~ borings in a metal as wll as the compression end temperature. block) . The hot wires form part of a wheat An example of this is shown in fig. 1. If stone bridge circuit and are kept hot by the NC-gasses are removed from the system the bridge current flow. When this is a it will operate as it was meant to. In o constant the temperature of the wires (~ ther words the effect is reversible. the resistence) will be determined by the The second effect may damage the valve conductivity of the gas in the chambers i.e. function. carrier gas and carriergas + sample compo If the NC-gasses are allowed to remain in nent. Since the carrier gas is choosen so the system for long periods it will result that its conductivity is different to the of in irreversible changes like thermal break gasses under analysis even small amounts con down of the oil around the valves and oxi sample in carrier gas will change its vol dative break down because of oxygen in the ductivity resulting in a difference of NC-gasses. tage between the hot wires. This voltage This problem is probably by far the worst difference is used as measuring reference chan arising from NC-gasses in the system. It and with the help of electronics the on a will show up mainly as leaky valves there ge is registered as a function of time by reducing compressor capacity, increasing normal recorder. in valve temperature etc. The detector signal is also sent to an form Even though the system is evacuated and re tegrator which convert it to diqital of the charged the compressor will have been dama enabling a calculation to be made ged to the extend where replacement is ne component concentration. a cessary. As will be seen, the sensitivity of such detector varies partly with the material measured and partly with the amount of the GASANALYSIS IN GENERAL material being measured. For this reason the detector must be calibrated. To be able to control the evacuation and This is done by analyzing a standard gas charging processes mentioned it is necessa mixture, the composition of which must of ry to have a method for determining the to course be known qualitativly and quantita tal content of NC-gas. tivly. For example air can be used with a The NC-gas content in refrigerants is tra volume lo-loo times smaller than the volume ditionally determined by suitably freezing used for the analysis itself. out or absorbing the refrigerant. Then the For each of the components in the standard to amount of gas residue can be determined by gas a factor is calculated corresponding a pressure - volume measurement. concentration divided by integrator output. the sensi For taking these types of measurements the This factor is an expression for and re has over the years been developed speci tivity of the detector to the component al glass equipment, measuring procedures concentration range concerned. been used etc. ( 2, 3, 4,) • For years gas chromatography has
481 by compressor producers in connection with sturb the system. compressor life time tests and associated If the gassample is transferred direct and analyses of ageing processes of insulating a suitably thin, diffusion thight, clean materials, oils, etc. (5,6). tube is used the necessary amount of samp However to the authors knowledge this tech le willbe so small that under no circum nique has not been used within refrigerator stances willconditions in the system be in production - perhaps tecause there are some fluenced. difficulties in the sampling procedure. Therefore it is not necessary to re-evacu ate and charqe the svstem. If the above-procedu~e cannot be carried REFPIGERATION SYSTEM CHARGE - GASANALYSIS out the sample must be'taken in a gasbomb. The size of the gasbomb and the procedure When analyzing the gas charge in a refrige involved is however of some importance. ration system the task should be regarded In our experience the most appropiate me as consisting of two operations: thod is to connect the born~ to the system (e.g. to the compressor process connection) 1. sampling so that it becomes part of the system vo 2. gas chromatography lume. In this way it is evacuated and char qed in the same way as the rest of the sy The results of the gas analysis will be in ~tem. 1<'7hen the sample is due tc be taken fluenced by the temperature and pressure in the bomb is merely removed from the system the refrigerator system at the time the gas and completely sealed, A suitabl~ volume sample is extracted. This is because Freon for the bomb will be 100-200 Ncm . can be part gas, part liquid and partly ab There will always be a certain error when s~rbed in the oil. Conditions for NC-gasses sampling by the bomb method.The magnitude are however a little different. They are of the error must be established by not soluble compar in oil when their partial pres ing the direct gasanalysis with analysis sure are sufficiently low e.g.(200 rnm Hg taken by bomb (both analyses on the same and their temperature is sufficiently high system). e.g.> 20°F (1), and they do not condense of course. Thus, the vol % of NC-gas found in a gas SA~PLE CHROMATOGRAPHY analysis will depend on how much Freon in gas form there was at the tirr.e when the Chromatography is carried sample out by taking was taken. from the bomb or refrigeration To determine system it the total content of NC-gasses self the amount of gas which in a refrigeration is necessary system therefor requires to rinse and charge the supply-lines analysis of and the composition of the gasphase the analysis loop in the chromatograph. and determination of how large the gasphase From the is. analysis loop - which i3 calibra ted accurately to contain one em at 760 mm Hg and 20°C for example - the gassample is lead to the column by SAMPLING PROCEDURE the carrier gas. In the column the separation occur as pre viously described, It is our experience that it is easiest to By using the factor arrived at earlier the take samples of the gasphase when the re concentration of the components found can frigeration system is pressure and tempera be calculated ture equalized. and recorded. The special conditions we work with This is done, after the for system has been in such analysis can be seen in fig. operation for the required 2. time, by stop As a result of the gasanalysis the ping the unit and opening content the cabinet door. of NC-gas = PCNC is found from the sample In our standard procedure a system is thus taken. pressure- and temperature equalized in a As NC-gasses are only found minimum of 12 hrs. in the gaspha se (provided PCNC .(.....,5 vel %) the next The gassample is sta extracted and at the same geis to determine the size of the gasphase time measurements are taken of equalized in the system concerned. pressure and temperature. O 482 the time of sa~pling can be calculated of the individual measurements. from an absorption diagram as shown on fig. The partial differentials are obtained by 3. From this diagram the wgt % oil in the the differentiation of eg. 3b: Freon + oil mixture under pressure and tem peraure equilibrium can be determined. ~ Ve12.· FC~C The amount of Freon absorbed is calculated ct;_ICI2. "' -jo, - R:::;~C.. thus: ~ " !t-o· GQ:i,.t ·V'el2· fcNC GR12 (oil) Cl w ( lo-o -PC I\,(.) . Vl/"1.. ?!~c.. . v2i2. .f'cNc . (1~~-o-wl and then the amount of Freon remaining in 7 the gasphase is calculated: 0 ~ " (I~ -:. PC. NC.) · W GR12(gas) GR12(total) - ( 2) dtvc, . [Ge~;_ ~ ~ 66d] · PC.NC dV£12- . 1~ -f'c.t..tc.. atvc [Ge1l- :,. ~ · s.o&J- VIC 12 · I~W = ( -1o-o -7 FCrvc.) 2. CALCULATION OF THE TOTAL AMOUNT OF NC-GAS altM:- IN A REFRIGEP~TION SYSTEM EXAMPLE OF CALCULATION How much vol % NC-gas there was in the gas phase of the system was determined by the A refrigeration system has the following gaschromatographic analysis of a sample of charges: the charge, i.e.: 185 g Freon 12 PCNC ( 3) 365 g oil measured Equalizing pressure and temperature are me- This equation expresses that the 2. val % NC-gas corresponds to the proportion asured as 3.7 kp/cm and 20°C. between the total amount of NC-gas and the From fig.3., the oil-Freon phase contains amount of Freon in the gasphase + the amount 70 wgt %, i.e. W = 70. of NC-gas, under standard conditions. 'A gassample is taken containing 2.0 vol % From this we obtain the total amount of NC NC-gas at 1 atm. and 20°C, i.e. PCNC = 2.0. gas as: The specific volume of Freon 12 is 0.20 1/q (see ref. 7) at l atm. and 20°C. i.e. VR12 G-Q-QCqa.o) • ve~2 · R::rvc NC (3a) : 0.201. ~o-o- PCIVC Then the absolut amount of NC-gas can be By using the obtained expression for deter~ined by eq. 3b: GR12(gas) eq. 3a becomes: NC 0.126 l. can also NC The absolute error on this value be determined, using the following factors: and the total amount of NC-gas is determi GR12 185 g S(GR12) =:s g :-!:2 ned. w 70 wgt % S(W) wgt % Gail 365 g s (Gail) :!5 g VR12 0.201 1/gS (VR12) =~0.001 1/g val % ESTIMATED ERROR ON NC-GAS DETERMIN~TION PCNC 2.0 val% S(PCNC) =!0.05 Pequa. 3.7 kp/cJi::o.l kp/c~ As can be seen from eq. 3b, the absolute Tequa. 20 "c :: o. 5 "c NC-gas content is a function of the follow ing 5 parameters: The partial differentials are calculated to: NC ~ f(GR12(total),W,Goil,VR12,PCNC) Ss.~lt 0.02.01./ l If the root of least squares principle is ~z;z . used to determine the absolute error of NC d~C.. s it becomes this: ow • w o.o!Qo2 t 2 2 c~ ) (~ c .•\ ONC.. $ . vt..e (()M: )2. (I . 0'00 Cf )._ (4) "0 ~c • sVJC.i 2 0. o-o-ob -t + ~2,· S~£tZ)t(~·W2. oVRI2. Eere, 483 The absolute error is calculated according not frightenigly to eq. 4: expensive. ONC 0. 065 l The relative error is determined: NOMENCLATURE d"Nc (relativ) "-50 %. NC (1) = Total volume of N~-gas in a refrigerator system Apart from the estimated error determinati PCNC (vol %) ~he vol % NC-gas on showing how large the error in a sam really is ple of the gasphase it also points out which of at stan- the individual dard conditions measurements produce the largest effect. GR12(total) Total Freon In this case it turns out charge that the weakest GR12 (gas) (g) : Amount point is the determination of Freon in the gas of the amount of phase Freon absorbed in the oil, and this result Gail (g) Amount of must be considered as oil charged the quite reasonably. compressor w (wgt %) "" ~Tgt % oil in the Freon - oi 1 CONCLUSION phase Pequa. (kp/cm) Equalized pressure Tequa. ( oc) Equalized temperature Overall, the following operations are ne VR12 (1/g) Specific volume cessary (after the rutine has "" of Freon at been establi gas analysis conditions shed) to determine the total content of NC 6Nc (1) Bstimated error for NC gas in a refrigerator system with the use of gas chromatography: l. Run the system in 2. Allow pressure and temperature to equ REFERENCES alize 12 hrs. min. with a open cabinet door 1. Baldwin,R.~., 3. take a sample and S.G.Daniel of the gasphase Journal of the 4. At the same time institute of Petroleum measure equalizing Vol. 39, 1953, pressure and temperature p.lOS. 5. Determine the content of NC-gas in the 2. sample McCloy,G.S., and A.W.Haley with a gas chromatograph Refrig.Engnq. 6. Calculate size of gasphase Bd. 39, 1940, p.l2. 7. Calculate total amount of NC-gas in the system 3. Parmelee,H.M. Refrig.Engng. We find that this procedure offers a series Bd. 59, 1951, p-573. of advantages which make it attractive to producers of units: 4. Steinle,H. K<etech. l. The analyses are reproducible - all Bd. 8, 1956, p.39. though it should be pointed out that each time a gassample is taken the to 5. tal amount Spaushus,H.O., and R.S.Olsen of NC-gas in the system Refrigerating becomes reduced. Engineering Febr. 1959, p.25. 2. The estimated error on the analyses can be easily determined. Calculations 6. Doderer,G.C., and H.O.Spauschus have shown that it is of a reasonably magnitude. ASHRAE Transactions vol.72, part II, 1966, p.IV.4.1. 3. Analyzing time is very short 5-15 min. At the same time need for manual ef 7. Thermodynamic forts is very small. properties of Refrigerants Publ. by ASHRAE 1969 4. The refrigeration system under analy Data for refrigerant 12 sis need not be removed from the pro duction line. There is a disadvantage. Knowledge of gas chromatography is a necessity and gas chro matographic equipment must be made avail able. This disadvantage must not however be exa gerated, gas chromatography is a very wide spread analyzing tool and the equipment in the simple form necessary to carry out the measurements we have been talking about is 484 ANALYSIS DETAILS: Gas chromatograph 5830A Hev.11ett Packard Column 2 m 1/8" steal tube ~so filled with Porepak Q Oven temperature 120°C Detector Hot wire 200°C Injection temp. 200°C Carrier gas He 12. -r flow 30 ml/min 2 "';yA~ pressure 4 kp/cm I i VT Analysis loop 0.5 c~ 10 q p Vo-f% Nz. 1.o ~.o y, ~. -··· -~-· -- 'l +q.3 485