GB2121818A - Lubrication of piston-type natural gas re-injection compressors - Google Patents

Abstract

A method of lubricating a piston- type natural gas re-injection compressor using as lubricant a polyglycol fluid having the formula RO-(CH2-CHR'-O)x-R'' where R and R'' are hydrogen or C1 to C4alkyl groups, R' is hydrogen or a C1 to C4alkyl group and x is at least 15.

Description

SPECIFICATION Lubrication of piston-type natural gas re-injection compressors It is necessary, in some oil fields, to compress surplus natural gas and to re-injectthis into the oil reservoir. The gas volumes may be very large (106 m3/day) and the injection pressures are high, e.g.

300-400 bars. The combination of the nature of the gas, the flow and the pressure has caused lubrication problems with reciprocating compressors leading to scuffing, and high rates of wear of piston rings and rod packings.

An investigation of the problem showedthatthere was an unexpected loss of lubricant into the high pressure gas. For instance, spindle oil exposed in an autoclave to methane at 300 bars and 120 C, distilled and condensed in the cooler parts of the autoclave; cylinder oil did not evaporate to such an extent, but there was some loss to the gas. To compensate for this loss, the lubricant supplyto the cylinders and packings had to be increased to at least ten times the manufacturers' recommended feed rates.

It has now surprisingly been found that these problems can be solved by using as lubricant a selected class of polyglycols.

Accordingly this invention relates to a method of lubricating a piston-type natural gas re-injection compressor using as lubricant a polyglycol fluid having the formula RO-(CH2-CH R'#O)x#R where R and R" are hydrogen or C1 to Cq a Ikyl groups, R' is hydrogen ora C1 to C4alkyl groupandx is at least 15 and preferably lower than 150, more preferablyx is 25-100.

From British patent application 2,075,544 it is known to lubricate piston-tapecompressorsforuse in the processing of natural gas with polyglycol of the above type wherein howeverx has much lower values, namely preferably 2 to 12. The problems involved in re-injecting natural gas into oil reservoirs are not touched upon andthe use of the present class of polyglycols is not mentioned.

The present polyglycols preferably is a copolymer of ethylene oxide and propylene oxide, preferably in a molar ratio of about 50/50, x being between 40 and 50.

Suitable molecularweights are between 1000 and 7000, preferably between 2000 and 2500, and suitable viscosities are between 100 and 2000 cS at380C, preferably between 125 and 250 cS at 385C.

Although additives are not essential in this oncethrough lubrication, they can be used, e.g. to improve storage and handling of the fluids. These additives should not cause well impairment (blocking of the porous rockstructure) having regard to the fact that some lubricant passes with the gas into the injection well.

It has beenfoundthatthe present neat lubricant has given none ofthewell-impairment problems experi ended with mineral lubricating oils.

This invention will be elucidated by the following experiments.

To determine the loss of lubricant more precisely, a small bomb was made, of 50 ml capacity, fitted with a light aluminium linerto contain 500 mg ofthe lubricant to be tested. Gas was taken from cylinders of 95% methane and compressed to 400 bars by a small oil-free compressor. The gas was led into the bomb bya 5 mm tube terminating above the oil surface, and outthrough a needle valve controlling the pressure to a meter. The gas flow was about 20 gimin, and it was allowed to flowforabout 10 minutes.

The effect of pressure itself would not be expected to cause evaporative loss of the oil sample: it must act through a change in the density of the high pressure gas. Experiments were therefore made to observe the effect of pressure (density) constant temperature, and of temperature at constant gas density. The resultsareshown in the Table.

At a given temperature the loss of spindle oil increases sharply with pressure of density, and it is very significant even at as low a temperature as 39 C.

There is no loss when nitrogen is substituted for methane. The temperature variance ofthe oil loss at constant gas density is very much less than that ofthe normalvapourpressure(1.6x 10#5mm Hg at400C,4x 10-3 mm Hg at 1000C).

The less volatile HVI Brightstock (normal vapour pressure6.4x 10-5 mm Hg at 1000C) shows only a small loss into high pressure methane.

Then the test was repeated using a lubricant according to this invention, namely a copolymer of ethylene oxide and propylene oxide, molar ratio 50/50, viscosity 210cS at 380C (lubricant A). Surpri singlythis lubricant showed no loss, except for 3% of light ends, which may be removed bytwo prior treatments with methane at 400 bars.

Evaporation of oil into high pressure methane

Temper- Pressure, Gas Oil Oil in ature, bars density, loss, gas, ppm OC g/ml mg by weight Spindle Oil 39 239 0.173 12 40 293 0.197 28 67 408 0.238 148 840 62 286 0.173 12 73 408 0.218 ' 152 730 100 276 0.144 25 120 340 0.172 89 360 408 0.194 340 1700 Brightstock 100 408 0.194 8.7 37 In addition to the absence of evaporation losses thus avoiding the increase of manufacturers' recommended feed rates, this lubricant greatly improved the effectiveness ofthe present compressors, giving a large reduction inthewear rates of pressure packings and piston trims, and consequently allowing much longer runs between maintenance shutdowns. Packing lives, one of the most critical factors, have for instance been increased from 600 hours (for a commercial compound mineral compressor lubricant) to 4000 hours (for the present lubricant A).

Claims (6)

1. A method of lubricating a piston-type natural gas re-injection compressor using as lubricant a polyglycolfluid having the formula RO~(CH=CHR'~ )x~R" where R and R" are hydrogen or C1 to C4 alkyl groups, R' is hydrogen or a C1 to C4 alkyl group and x is at least 15.

2. A method according to claim 1, in which x ofthe polyglycol islowerthan 150.

3. A method according to claim 1 or 2, in which x ofthe polyglycol is between 25 and 100.

4. A method according to any one of claims 1-3, in which the polyglycol is a copolymer of ethylene oxide and propylene oxide.

5. A method according to claim 4, in which the molar ratio of ethylene oxide and propylene oxide of the polyglycol is about 50/50.

6. A method according to claim 1 substantially as hereinbefore described.