cONF- 7&04/16- -9 OPERATION OF MAMMOTH PACIFIC'S MP1-100 TURBINE WITH METASTABLE, SUPERSATURATED EXPANSIONS

Idaho National Engineering Laboratory

ABSTRACT 0 s-v I INEL's Heat Cycle Research project is currently examining the potential improvements in pertbr- mance achieved when metastable, supersaturated vapor expansions are allowed in the binary cycle turbine. During these turbine expansions, the working fluid expands into the equilibrium two- phase region. Ifthe fluid is brought to equilibrbm at that point, liquid condensate would be present. The formation of the c o n k t e is not an instants- nmus process; it requires the chance grouping of molecules or it requires nucleation sites for the drops to fotm. Because of the delay in umdensate formation, the vapor is referred to as supersatu- rated. The turbine expansion is considered to be a metastable process.

Expansion of a vapor into the equilj%rium, two- phase region in a turbine is not unique. In con- densing steam turbines, the W h n Line estimates the extent to which the expansion can proceed while m a i n t a i i a supersaturated vapor (no con- densation). The co~densation behavior of the hy-

for binary cycles is not Srpectedtobethesameasthatofsteam. An isobutane working fluid or a mixture of isobutane and a heavier, minor component provides the ' highest cycle perfbrmance for the resource tem- peratures of interest Isobutane, as well as the

The Idaho National Ebgineering Laboratory's Heat Cycle Research project continueS to develop a technology base that will permit increased use of moderate-temperaturehydrothexmalresourcesto generate electrical power. Project investigations have confinned the v i a b w oftechnologies that

binary power cycle perfinmame to ap-

of the coflcepts under investigation is the use of metastable, supersaturatedturbme expansions. These expansions support a supersaturated work- ing fluid vapor; at e c l u i l j i conditions, liquid condensate wodd be present during the turbine expansion process. Studies suggest that ifthese expansions do not adversely affect the turbine per formance, up to 8 to 10% more power c d d be produced from a given geothermal fluid Deter- mining the impact ofthese expansions on turbine performance is the focus of the project investiga- tiom being reported. This work is supported by the U.S. Department of tary for Enerpy Efficiency

thermodynamicmaximums. One

inFigure 1. Inthis cycle,

its critical pressure. In a con-

engintxringtechnologies necessaryto mcorporate these concepts into the design and operation of a b W power Plant.

before entering the turbine to ensure that the ex- pansion process in the turbine (represented in Fig- ure 1 by point 3 to point 4) will remain completely

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency Thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

DISCLAIMER Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.

, ., > ' . . * ..

outside the *phase region. Demuth4 suggests that a modified cycle where the ideal turbine ex- pansion process passes through the two-phase re- gion ( r e p r e s d by the process from point 3' to point 5) could proceed without adversely affecting turbine performance. Mer a theoretical d- nation ofthe um&nsation behavior of the hydro- carbon working fluids m these expansions, De- muth concludes that droplets might not form. droplets did fimn, they initially would be vexy small and tend to evaporate as the expansion pro- Ceeded

3

esupefheatlng (4,q

L W o r k l n g fluid pumping (13

, INEL 2 1875 S

Figure 1: Binary Cycle Showing Two Types of Turbine Expansions

During the metastable expansions in a steam tur- bine, the mephase region is entered near the end of the expansion process (the Einal turbine stage). In contrast, the proposed metastable expansions with the isobutane working fluid enters the 6m phase region much m e r in the expansion proc- ess (see Figure 1). Depeding upon if and when

the vapor as it enters the turbine rotor. Ifthe droplets are present at the rotor inlet, there is a higher potential for degradatim in performance and erosion of the rotor and mes (nozzles).

If it could be confirmed that these expansions did not impact turbine perfonaance, an additional cy- cle performance improvement of up to 10% could be realized.

condensationoccurs,dropletsmaybeentrainedin

HCRF FED INVESTIGATIONS

To determine the impact of these expansions on the perfonnaace of a binary cycle turbine, the pro- ject conducted a Series offield investigations at the INEL's Heat Cycle Research Facility (HCRF) located in C a l i f d s Imperial Valley. During the first phase of these investigations, the conden- sation behavior of these metastable expansions was examined with a cmverging-diverging nozzle. In the second phase of these investigationS, tur- bines were tested to determine the degree to which the expansions could enter the two-phase region (the level of supersaturation) before affecting the turbine perfarmance.

The two-dimensional, converging-diverging nozzle expansions simulated an isentropic process in a turbine. Diffetent nozzle inlet conditions were examined to determine those conditions that re- sulted m condensate forming during the expansion A window on the nozzle allowed the expansion process to be monitored. A laser-based droplet- detection system was used to identify the onset of condensate formation and to size the droplets that

fim 550 to 650 psia (above the isobutane critical pressure). By changing the nozzle inlet tempera- ture at a fixed inlet pressure, the degree of super- saturation in the expansion was varied. Testing was conducted with an isobutane working fluid, as well as an isobutane-hexane mixture.

he nozzle inlet pressure was varied

The nozzle testing confirmed that these metastable expansions initially support a supersaturated va-

sible to delay the onset of condensate fbrmation m the node until a maximum equiliium moisture level of 6 to 7% was reached (at equiliibrium, 6 to 7% of the fluid would be liquid). At the onset of condensate formation, the droplets initially e v a p rated as the fluid expanded. As the inlet condi- tionswerevariedtofurthermcreasetheequil.ii rim moisture level, the con-te that formed did not evaporate. The laser, dropletdetection system proved to be very effective in identifykg the onset of condensate hrmation and its location along the expansion path. The results of the test-

por without condensate fOrmatioa4' It was pos-

* .. . , - 4 ' .,: - I

i t l '

ing to size the droplets that f m e d were inconclu- pansions m any erosion damage to inter- \ Sive. rial components.

During the second phase of the HCRF investiga- tions, an axial-flow, impulse turbine and a radial- inflow, reaction turbine were installed and tested Barber-Nichols, Inc., provided the axial-flow tur- bine, generator, load bank, and associated control equipmat. The turbine had a nominal power out- put of -40 hor~epwtx. RstonoW Corporation provided the radial-inflow turbine. In conducting the turbine tests, the inlet and exhaust pressures were constant test conditions. The turbne inlet temperature was varied to obtain the desired levels of supersaturation or equiliium moisture levels in the expansion process. The turbine tests were PrimariIy conducted at an inlet pressure of600 psia with an isobutane working fluid. Limited testing was also conducted at an inlet pressure of 550 psia, as well as with an isobutane-hexane mixture.

The testing with the turbines performance was not impacted at inlet conditions supporting a supersaturated vapor in the n o d e mvestigatiions.' The impulse turbine performance was not affkted until the conditions leaving the turbine (actual) were within the two-phase region (the expansion would have a maximum equilib- rium moisture content of -25%). The radial-h- flow turbine perfonnane was not impacted until

MAMMOTH INVESTIGATIONS

At the conclusion of the HCRF hvestigations, investigators had shown that the metastable ex- pansions could proceed without adversely a f k t - ing turbine @brmance. However, hestigators had not resolved whether the pedmnance could be maintained over the expected Iifk of a com- mercial turbine. (Over time, the presence of con- densate could damage the component surfaces ex- posed to the expanding vapor.) With the closure of the HCRF, it became necessary to 6m-l an al- ternatefacititywhereaturbineddbeoperated with the metastable expansions for an extended period. With the assistance of Rotoflow Corpora- tion and C.E. Holt Company, the project reached an agreement with Mammoth Pacific Limited Partnership (MPLP) to continue the metastable expansion investigations at one of MPLP's corn.. mercial hdities near Mammoth Lakes, Califor- nia. At this location, MPLP operates three power plants, with a total design capacity of40 me). All three are binary plants using an isobutane working fluid MPLP proposed conducting this investigation at the Unit 100 of the MP1 plant . (Mp 1- 100).

As part of the agreement with MPLP, the project

eumtentof-13%. At

conclusion of the

turbines was inconclusive as to whether the ex- and operated with a subcritical turbine inlet pres-

sure. The Mammoth kilities produce electrical power for commercial sale; they are not designed for research adivities.

After evaluating the performance ofthe MP1-100 facility over a range of different operating pa- rameters, the project concluded that it would be possible to operate the fiicility with the metastable expansions, though the inlet pressure would be below the critical pressure of the isobutane work- ing fluid A mutual decision was made to con- tinue with extended operation investigation, with the following conditions:

The new turbine rotor and set of vanes were to be the same as the existing design. The operation with the metastable expansions would have minimal impact on the power p m duced by MPLP's facilities at Mammoth The turbine would operate at as high an inlet pressure as possible. The turbine would operate with 1 to 2°F of superheat at the inlet. During a b n d periods of operation, MPLP could adjust the turbine inlet ConditioIlS as it cansidered necessary. The turbine could be operated at the project's requested conditions during the period be- tween October 1995 and June 1996. At the end of May 1996, a mutual decision would be made whether to continue the investigations.

The new turbine components were obtained fim Rotoflow and installed in November 1995. Before installation, the component surEaces that would be exposed to the expanding vapor were photo- graphed to establish the pretest condition. After installation of the new components, the turbine operated for -140 hours at the nominal conditions to establish a baseline performance with com- pletely dry expansions. During this baseline pe- riod, the turbine efticiency varied between 70 and 85% with changes in the ambient air temperature. The brine utilization varied from -5.7 w-hr/lbd (watt-hr per pound of geothermal fluid) at an air temperature of 32°F to -3.2 w-hdlbd at 64T.

The baseline period was used to define the inlet pressure condition for the operation with the meta- stable expansions. This was accomplished by conducting a series of tests where the turbine vanes were sequentially closed in 5% increments, throttling flow, and raising the turbine inlet pres- sure. The e&ct of throtthg with the vanes on the turbine inlet pressure and turbine efticiency is shown in Figure 2. At the nominal turbine operat- ing conditions, the turbine inlet pressure WAS -275 psia. The desired minimum pressme for the ex- tended period of operation was 450 psia. To achieve this pressure, the vanes would have to be closed to between 40 and 50% open. Although this resulted in a significant degradation in the turbine &ciency, the brine utilization during this series of tests did not change significantly. The decrease in efliciencywas ofk t by an increase in the isentropic enthalpy change (the potential work across the turbine) and a decrease m the brine flow requirement. Since excess brine &mMPl- 100 could be diverted to one of the other fkcilities and increase its power pro&ction, it appeared that the pressure of 450 psia could be attained without adversely affecting the total power output fiom MPLP'S facilities.

90 500

85 450

- 5 80 400 $ V a

ID - s u

350

70 300 c

5 75 0,

Q L

65 250 g S c

60 200 40% 60% 80% 100%

vane position,%epen

Figure 2: Effect of Vane Position on MP1-100 Turbine Perfibmince

The range ofturbine efficiencies m Figure 2 for a given fixed vane position reflects the variation in the measured isobutane flow rates. The turbine efficiencies presented in this paper are the actual turbine work divided by the isentropic enthalpy change between the inlet conditions and the ex-

8

h a w pressmi Except where indicated, 1 astheamb temperature changed. The turbine work is obtamed by dividing the power, with adjustments for generator ine&ciency and gearbox losses, by the measured isobutane flowrate. AtMP1-100, therearevariations in.the m d isobutane flow rate of-5 to lo??. Be- cause there are not similar variations m the other measured parametgs m the isobutane system, it is suspected that fbe fluctuations in the measured flow rate are inherent to the measurement and are not actually occurring. To &et the fluctuation in the m e a s d flow, the d u e recorded hourly was an average ofmu

OnNovember 13 tions at MP1-100 were adjusted to a pressure of 4 5 0 psia and a superheat level to 2°F. These conditions were a#ained at a turbine vane position of 4 5 % open. At these inlet conditions, the maximum e q u i l i i moisture duriugthe turbine expansion was -1 to 2%. During the operation of MP1-100 at these inlet d o n s , theturbine vanes were maintained at this brine flow rate held at a amstant isobutane flow rate was used to a superheat at the turbine inlet Turbine perfbrm-

plant (in terms ofthe amount of power produced from the brine flow utilized) was not affected by operating at the modified inlet conditions. During the first 100 hours of operation, the brine utilization varied f?om -5.9 w-hr/lb@ at an ak temperature of32"F to -3.6 w- hrhgat a temperature of 64OF. These d u e s were similar to those at the baseline conditions.

The MP1-100 turbine bas *ted with the modified inlet conditions since November. The turbine inlet conditions during this period are shown inbothFigum 4 and 5. Figure4 is a tem- peratumentropy (T-S) plot showing the nominal range of turbine inlet conditions (befii operation with the metastable expansions), a daily operating point during the metastable expansion investiga- tion, and the saturation curve for the isobutane working fluid. The turbines at the Mammoth fa- cilities are nominally operated at inlet conditions that are approximately at an entropy of S=l .190 btu/lb-R (as defined by the NISTl2 property codes). The operation at these inlet conditions ensures the expansions occur ccmpletely outside the two-phase region, as well as provide s&cient

of time aver the period the turbine has

cated in this figure, the inlet conditions were re-

vised after -1270 hours. On January 5,1996, the turbine inlet pressure was increased to -465 psia with inlet superheat levels of lo+0.5V. Before this rec;ision ofthe inlet conditions, there was a trend toward an increasing turbine inlet entropy (decreasing level of supersaturation). [At inlet entropies above the maximum dew point entropy (S=l.l806withtheNISTcode),theturbineex- pansion occurs completely outside of the two- phase region.] The turbine inlet entropy increased because the turbine inlet pressure had d e c r e a s e d

at -2°F). (the inlet superheat was mmtamed Changes in pressure were noted after periods of atypical operation in responseto activities at Mammoth's other plants or to curtailments in power production imposed by the utility. Al- though the turbine vanes were returned to the mdi- cated value of45% open after these periods of abnormal operation, the inlet pressure would be at different values, dqmding upon whether the vanes were opened or closed to reach the set value.

. .

1.19

1.185

k 1.18

2,

fc. 5

1 I 7 5

1.17 0 500 lo00 1500 200 2m 3Ooo

run Utne, hours

Figure 5: IW1-100 Turbine Inlet Conditions during Operation with Metastable Expansions

When the turbine inlet conditions were revised m January, the mode of operation was also changed. Along with reducing the amount of superheat at the inlq MPLP was asked to use the turbine vanes to maintain the inlet pressure at 465 +5 psia. The turbine was operated with the total

s

isobutane flow rate available (maximum pump capacity); small adjustments were made with the brine flow to adjust the level of superheat. As indicated in Figure 5, the turbine inlet entropy (and the level of supersaturation) has remained relatively constant since January. At the revised inlet conditions, the equilibrium moisture level in the expanding vapor is -3 to 4%. Based upon the nozzle investigaton where the e q u i l i i mois- ture level reached 6 to 7% before condensate formed, the MPl-100 turbine is being operated at conditions supporting a supersaturated vapor.

The performance oftheMP1-1OOturbinethrough mid-March of 1996 is shown in Figure 6. Effi- ciency detemhed from both the generator power measurement and the measured turbine inlet and exhaust pressures and temperatures are plotted with time. The efficiencies determined by both methods are in agreement The data indicate that there has been no degradation in turbine perform- ance with time while operating with these meta- stable expansions. The modification ofthe turbine inlet conditions in Janzay resuIted in no appre- ciable change in the turbine performance.

I00

95

90

55

50 N-5 D~c-95 Jan-gG Few6 Mar-96

time

Figure 6: MP1-100 Turbine Efficiency during Operation with Metastable Expansions

The observed variation in the turbine efficiency with the ambient air temperature (condensing/ex- haust pressure), raised some question that the cooter ambient air temperatures during the winter may have masked a degradation in the turbine per- formance. In Figure 7, the turbine perfonnance is

L

plotted as a fitnetion of the ambient air tempera- 8 ture for each month, as well as duringthepgiod

7

' 6

when the baseline perfixmame was established. These data show there has been no -tion in

temperatures, the turbine efficiency in March was

c

5

performance through mid-March. At similar air s essentially the same as it was in November at the start of the investigation.

30 40 50 60 70

ambient air temperature, F

Ambient Conditions on MP1- 100 Plant Performance

SUMMARY

The turbine at MPWs MP1-100 plant bas oper- ated since November 1995 at inlet conditions sup- porting a supersaturated vapor (as indicated by the HCRF ride tests) without any measured degradation in perf'ce with time. To operate the turbine at inlet pressures that would produce 100 Turbine Performance

stable expansions, it was necessary the hbutme flow with the turbine

A &gradation m turbine efficiency was associated with k e g with the ma, iadependent of operation with the metastable expansions. Despite the decrease inturbine efficiency, there was not a corresponding degradation cycle pertbtmaace (in - ofthe pawg p r h c d fimagiven flaw of brine). Ifthe rurbine-aJxi- h fbeen sized to maintain efIiciency at the reduced iso- butane flows, the cycle Performance codd have increased by up to -10% at the turbine inlet

providing le expansions.

the investigation with the metastable expansions have minimal impact on the power production. Figure 8, the brine utilization for theMP1-100 facility is plotted as a hction of the air tempera- ture for the cent periods of operation during this investigation. The data show that possible to maiatain the brine similar to the baseline cycleperfbmmc metastable expansions, even though the efliciency was 6 to 9 The data in Figure 8 show that the brine utiliza- tion improved during the January to March period Although this improvement is red, it cannot be at- triiuted to the operation with the metastable ex-

figuration of the flow through the heat exchangers that was mde inFebruary 1996.

The coIlstFafnts * imposedbytheikc equipment configmtion and the need to minimize

pansioas. It is the result of a more optimum con- onplantperfo~cehasdictatedthat beopgateda*~fiti=l~etP-

s u r a with -1OF of superheat at its inlet. These constraints have limited the level of supersatura- tion attained in the vapor as it is expanded through the turbine. As the ambient air temperatures in-

crease (raising the condensing pressure) m the spring, it may be possible to further raise the tur- bine inlet pressure without additional throttling with the turbine vanes. This may allow further modification ofthe inlet conditions to increase the level of supersaturatian m the turbme expansions and not adversely affect IvPLP's power produc- tion.

The MP1-100 turbine will continue to be operated with the metastable expansions through May 1996. The agrement between INEL and MPLP provides for the suspension or termination ofthe investigation then at the request of either party. (This provision provides MPLP the flexiiility to adjust its operations as necessary dwingthe summer when maximum revenues are obtained for the power produced.) Ifthe operation of the turbine with the expansions proceeds without in- terruption, MPLP will continue to provide the project operating data. Ifthe test is suspended, it is anticipated that it will resume in the fhll of 1996. With the decision to texminate the investi- gation, a nondestructive examination of the tur- bine rotor and vanes will be made. This examina- tion will be made at the first shutdown ofthe tur- bine for maintenance fillowing the decision to conclude the test. In the event of a turbine fkilure, the turbine componenb will be moved and ex- amined to evaluate the contriiution of the opera- tion with the metastable expansions to the fitihre.

Ifthe MPl-100 turbine continues to operate with no degradation in perfbrmance though the end of May 1996, the project will explore the possibiity of expanding these investigations to other hcili- ties.

REFERENCES

1. O.J. Dmuth, Analysis of Mixed @&warbon Binary Themtodynamic Cjclesfbr Mmkmte Temperature Geothermal Resources, EGG-GTH- 5753, February 1981.

sources' Using Regenerative Techniques, EGG- GTH-5720, November 1981.

3. C.J. Bliem and G.L. Mines, AhancedBinary Geothermal Power Plants Limits of Pei$onnance, EGG-EP-9207, Jan~ary 199 1.

4. O.J. Dmutb, Preliminary Assessment of Condensation Behavior for Hjldz.ocarbon EjEpan- sions Which Cross the Saturation Line Near the Critical Point, EGG-GTH-5960, July 1982.

5. D.Y. Goswami, S. Hingorani and G.L. Mines, "A Laser Based Technique for Particle Sizing to Study Two Phase Expansion m Tur- bines," ASME paper reprinted &om Solar Engi- neering, BookNo. H00630,1991.

6. sation Behavior of Supersaturated Turbine Ex- pansions," Proceedings of DOE Geothermal Pro- gram Review m, Berkeley CaHornia, April 1993.

G.L. Mines, "Investigations of the Conden-

7. the Impact of Supersaturated Vapor Expansions on Turbine Performance," Proceedings of DOE Geothermal Program Review XU, San Francisco California, April 1994.

G.L. Mines, "Field Investigations Examining

2. O.J. Demuth and RJ. Kochan, Analysis of Mixed Xy&marbon Binary Thermodynamic Cy- cles for Moderate Temperature Geothermal Re-