《流体机械外文翻译外文文献英文文献减轻高压注气压缩机爆炸风险.doc》由会员分享，可在线阅读，更多相关《流体机械外文翻译外文文献英文文献减轻高压注气压缩机爆炸风险.doc（16页珍藏版）》请在三一办公上搜索。
1、原文Mitigating Explosion Risks in High Pressure Air Injection CompressorsAbstractThis paper describes research undertaken by Encore Acquisition Company and the University of Calgary regarding the flammability safety of synthetic lubricants used in Encores compressors in their high pressure air injecti
2、on (HPAI) projects in Southeastern Montana. With over 2,270 e3m (ST)/d (80MMscfd) of installed air compression capacity discharging at pressures of 31.0 to 34.5 MPa (4,500 to 5,000 psi), a critical aspect of this project is the safe and uninterrupted operation of the compressors. Experience gained b
3、y Encore and other HPAI operators shows that the reaction of high pressure air with compressor lubricants in high temperature interstage and discharge regions of the compressors can be a source of trouble (destructive overpressures), even when synthetic diester-based lubricants are employed.Using an
4、 Accelerating Rate Calorimeter (ARC), samples of fresh and used synthetic lubricants were heated in the presence of air at initial pressures up to 34.5 MPa (5,000 psi). Self-heating rates and pressure responses were measured.The results highlighted the significant effect of pressure on auto-ignition
5、 temperature. Most significantly, the auto-ignition temperature of the diester-based lubricant dropped from the manufacturers reported level of 410?C (770?F) at atmospheric pressure to 180?C (365?F) at pressures in the range of 17.2 to 34.5 MPa (2,500 to 5,000 psi). Also, the auto-ignition temperatu
6、re of used (oxidized) synthetic lubricant was further reduced to values close to the operating temperature levels of the compressors. Finally, it was noted that the auto-ignition temperatures for different brands of diester-based lubricants were all very similar.The significance of this study is not
7、 only in the temperature data, but also in the discussion of several significant changes that Encore made to the design and operation of their high pressure air compressors as a result of this study. This information will assist future HPAI operators in designing safe and reliable air compression sy
8、stems.IntroductionImproved oil recovery of conventional light oils by high pressure air injection (HPAI) is becoming a well known process. With increasing oil demand and dwindling primary and secondary production-based reserves, more producers are showing increasing interest in HPAI. Typical example
9、s of such increasing interest include Encores eight injector 17 e3m (ST)/d (6 MMscfd) HPAI project initiated in 2002 in their Pennel Unit that has since expanded to 850 e3m (ST)/d (30 MMscfd) with an ultimate design of 1,700 e3m (ST)/d (60 MMscfd). Air has been continually injected at 34.5 MPa (5,00
10、0 psi) for four years in the original portion of the flood. In addition, a new 566 e3m3(ST)/d (20 MMscfd)HPAI project was initiated in the Cedar Creek-Little Beaver East Field located on the Montana/North Dakota border in 2004. This project has 28 injection wells and is operated at 31.0 MPa (4,500ps
11、i). Figure 1 is a map indicating Encores presence in the Cedar Creek Anticline over 160 km (100 miles) long showing the HPAI and waterflood projects sites. Safe and uninterrupted compressor operation is the most important factor in the successful implementation of every HPAI project. To improve oper
12、ational safety and reliability, ester-based synthetic lubricating oils are widely used in HPAI compressors. The advantages of these synthetic lubricants over petroleum-based lube oils include higher flash point, higher auto-ignition temperature(1, 2)(AIT) and higher detergency (low residue) . These
13、properties significantly minimize explosion risks during compressor operation. The standard test manufacturers use to measure the auto-ig(3)nition temperatures of these lube oils, ASTM E-659 and ASTM (4)D-2155 , are performed at atmospheric pressure. The flammability ranges of hydrocarbon fuels are
14、known to widen with an (5)increase in pressure . Thus, operating below the manufacturers recommended AIT would not eliminate the potential risk for an explosive reaction between the air and the compressor lube oil. This was thought to occur in an Encore Acquisitions HPAI facility in Southeastern Mon
15、tana. The objective of this work was to investigate the oxidation behaviour of fresh and used (partially oxidized) synthetic ester-based lube oils, Anderol 750 and Anderol 555, at different reaction pressures. The results were compared to those of other ester-based synthetic lube oils; namely, Shell
16、 Corena DE 150 and a custom formulated ester-based synthetic oil from Summit Lubricants. The safety improvements Encore implemented based on the results of this investigation are also discussed.Applicability of ARC tests to lube oilThe accelerating rate calorimeter was originally developed for the e
17、valuation of thermal hazard of substances. Detailed description of the ARC apparatus, its theory, experimental procedure and (6-8)data analysis are available in the literature .The classical analysis of ARC test data is based on a simple singular decomposition reaction of the form A Products, with t
18、he basic assumption that the reaction goes to completion, and the products do not interfere with the reaction mechanism. If the above exothermic reaction occurs adiabatically, all the reaction heat will be used in raising the products temperature. The resulting selfheat rate can be approximated by t
19、he following(6) : （1） The above equation can be rearranged to the form below (with the pseudo-rate constant ). （2）At the correct reaction order, k * has the same functional dependence on temperature as k*. Hence, experimental thermal decomposition data is used to obtain k* and a graphical iteration
20、is used on the Arrhenius model below to obtain the best fit for the reaction order and the kinetic parameters(6) . （3）Unlike the singular decomposition reaction on which ARC theory is based, formulated synthetic ester-based lube oils often contain various additives to improve their oxidative stabili
21、ty(9.10). The oxidation reactions of these lube oils and other complex mixtures, such as crude oils, seldom go to completion during an ARC test either due to oxygen availability or formation of refractory residual material, even in the presence of excess air. In addition, the observed experimental d
22、ata such as dT/dt and T need F to be corrected for the thermal inertia of the reaction cell. While self-heating rates upwards of 1,000?C/min are recorded, experimental heater power supply data verified that the ARC is not able to closely track actual self-heating rates in excess of 15 to 18?C/minute
23、. Therefore, measured maximum heating rates of greater than the 15 to 18?C/minute range should be viewed as qualitative indications of elevated oxidation rates rather than as the absolute SHR values. In spite of these limitations, ARC tests have provided useful values for the kinetics parameters for
24、 even complex mixtures such as crude oil(11) .Experimental ProcedureMaterials The synthetic ester-based lube oils Anderol 750 and Anderol 555 used in this work are of ISO viscosity grade of 150 and 100 , respectively, supplied by Encore. To investigate the effect of inservice oil use on the potentia
25、l risk of auto-ignition, both fresh and used (partially oxidized) oils were tested. The used oils were sampled from different points along the compressor train. Table 1 summarizes the oil type, location description and the ARC test number in which the oils were used. The Anderol 750 sample used in T
26、est 3 and Test 4 was taken from the final stage scrubber which is after the pulsation bottle and the cooler, while the one used in Test 5 and Test 6 was from the pulsation bottle directly after the final compression stage. Thus, the former Anderol 750 sample had been subjected to a high temperature
27、and pressure for a longer duration. The Anderol 555 used in Tests 9 and 10 was obtained from the last compression stage with a normal operating temperature and pressure of about 149?C (300?F) and 31.0 MPag (4,500 psig). The Shell Corena DE 150 is a 150-ISO viscosity grade oil consisting of about 90
28、99 mass% synthetic esters, and 1 10 mass% proprietary polymer additives . The Summit oil is a custom formulated ester-based synthetic lube oil for high pressure air compressors.ARC test Procedure The ARC tests were performed in a modified CSI-ARC apparatus. Briefly, the ARC consists of the calorimet
29、er unit with the sample-holder (see Figure 2), power supply, temperature control and main processor units. The main processor is used for setting the run conditions and controlling the experiment. It also processes the results to obtain kinetic and thermal parameters from the experimental data. The
30、two ARCs at the University of Calgary have been extensively and specially modified for reliable operation up to 41.4MPa (6,000 psi), while tracking highly exothermic combustion reactions. They use specially developed 9 cm reaction cells made of Hastelloy-C weighing approximately 24.3 gm. The instrum
31、ent is calibrated and drift-checked before each run. In each test, about 0.2 g (Table 1) of the synthetic lube oil sample was loaded in the reaction cell and mounted in the ARC heating assembly, as shown in Figure 2. The ARC tests were performed in a closed mode in which the oil-loaded reaction cell
32、 was filled with air to the desired pressure and sealed. No mass exchange occurs across the cell boundary during the test; the cell boundary includes the volume of tubing connection to the pressure gauge. The ARC system was heated to the starting temperature of 50?C and operated in a heat-wait-searc
33、h (HWS) mode. In the HWS operating mode, the sample was subjected to repeated cycles consisting of heating in steps of 5?C, followed by a 20 minute wait period (for thermal equilibration between the lube oil sample and the reaction cell) and then a search for self-heating. During the latter, the mic
34、roprocessor searches for a self-heat rate (SHR) greater than a preset value of 0.025?C/min. Once an exotherm is detected, the exothermic (oil oxidation) reaction is allowed to proceed adiabatically. Four test pressures were selected in order to investigate the effect of pressure and to reflect the i
35、nterstage pressures in the real operation.Results and Discussion temperature and Pressure Profiles The ARC test conditions used for the two synthetic lube oils, Anderol 750 and Anderol 555, are summarized in Table 1. Typical accelerating rate calorimeter responses for the HWS test method are shown i
36、n Figures 3 and 4 for fresh Anderol 750 tested at 17.2 MPag (2,500 psig) and at 34.5 MPag (5,000 psig), respectively. The temperature profiles show the ramp of stepwise temperature increase due to the HWS scheme with the occurrence of an exothermic reaction (self-heat) beginning after about 950 min
37、at 180?C (Figure 3). No additional exotherms were detected in the subsequent HWS steps until the maximum operating temperature of the system (500?C) was attained and the system automatically shuts down resulting in the observed rapid cooling. The pressure profile closely followed the temperature pro
38、file. The rates of change in temperature and pressure during the exotherm are shown as insets in the above figures. At 34.5 MPag (5,000 psig) test pressure(Figure 4), the major exotherm occurred earlier but at the same starting temperature as the 17.2 MPag (2,500 psig) test (850 min, 180?C). Based o
39、n the temperature vs. time responses as shown in Figures 3 and 4, the self-heat rates due to the exothermic reactions of the lube oil and air at different initial pressures are presented in Figures 5 and 6 for the fresh and the used Anderol 750, respectively. Similar SHR plots for fresh and used And
40、erol 555 are shown in Figure 7, while Figure 8 shows the exothermic self-heating of the Corena DE 150 oil and the custom formulated Summit lubricant. The regions with no data points in the above figures indicate the absence of any detectable exothermic activity in those temperature ranges.Self-heat
41、Rates and AItThe SHR plots in Figures 5 and 6 show a characteristic ramp-up in exothermic activity at the lower temperature end of the data (180195?C and 155 165?C for the fresh and the used Anderol 750, respectively) followed by two distinct pressure dependent behaviours. For the 17.2 MPag (2,500 p
42、sig) and the 34.5 MPag (5,000 psig) runs, the initial portion of the self-heat region is followed by an essentially instantaneous 20-330 fold increase in the SHR, and an eventual decline in the latter. Following the initial SHR phase, the runs at 500 psig and 1,000 psig show phases of increasing SHR
43、 albeit much lower than the ones at 17.2 and 34.5 MPag (2,500 and 5,000 psig) followed by the final decline in the SHR.The self-heat rates are relatively insensitive to the test pressure for self-heat rates of less than 0.3?C/minute. The initial energy generation is believed to be associated with li
44、quid phase reactions for which there is an excess of oxygen and reactive hydrocarbon in the liquid phase. Once the self-heat rates approach a certain level, the liquid phase reactions become controlled by the rate of transfer of oxygen (i.e. on the oxygen partial pressure in the vapour phase and mas
45、s transfer resistance), as well as on the concentration of reactive hydrocarbon in the liquid phase. In the absence of the onset of a reaction zone involving the vapour phase, the oxidation reactions responsible for the energy generation will be associated with liquid phase reactions and will termin
46、ate when all of the reactive liquid phase hydrocarbon fractions are consumed. This appears to be the case for the 3.4 and 6.9 MPag (500 and 1,000 psig) tests.A comparison of Figures 5 and 6 reveal that under similar test conditions, the oxidized oil starts to react earlier, and at a lower temperatur
47、e than the fresh oil. This is believed to be due to the formation of liquid phase oxidized species under the in-service conditions of the commercial compressors. The oxidized species (9)participate in a number of liquid phase free radical reactions . The above observations are also true for Anderol
48、555 (Figure 7).The shape of the self-heat rate curves is indicative of different mechanisms controlling the overall heat generation rate at a given temperature. At the lower temperature regions, the SHR traces exhibit a behaviour which can be approximated using an Arrhenius-type expression which imp
49、lies a semi-logarithmic relationship between self-heat rate and inverse absolute temperature. This behaviour will appear as a straight line on the self-heat rate plots shown in Figures 5 to 8, inclusive. Over other temperature ranges, the self-heat rates are relatively insensitive to temperature which suggests that mass transfer to, and within the liquid/solid phase, is dominating the self-heating rate.Figure 8 s
