1

TRNSYS Modeling of 1 MWe Saguaro Plant

Greg Kolb Sandia National Laboratories

NREL Vahab Hassani

Why TRNSYS models of Saguaro?

Why TRNSYS models of Saguaro?

• SunLab will help APS assess the performance of Saguaro

• Should Saguaro pursue storage?

• Existing trough models (EXCELERGY, LUZERGY) are inadequate

– Do not model constant flow mode (Saguaro) – Do not model thermocline interactions with rest of plant

• Existing thermocline models (Pacheco, NEXANT) are inadequate

– Do not include thermal losses or axial conduction – EXCEL based models run slowly and make annual simulations impractical

• Initial TRNSYS models of Saguaro and future Saguaro are now available

– Constant flow, no storage (Saguaro) – Constant temperature, with storage (future Saguaro) – An annual simulation only takes 1 or 2 minutes

2

Saguaro(constant flow, no storage)

Saguaro (constant flow, no storage)

New Type 297 (const flow) created from STEC Type 197 (const T)

Temp/flow is coupled between solar and ORC

Curve fit model of ORMAT data at constant flow:

Also, delay ORC startup after achieving startup condition. Output kWe = f(In_temp), Out_Temp = f(In_temp). 3

Initial Saguaro TRNSYS Simulation Insights

Initial Saguaro TRNSYS Simulation Insights

• TRNSYS predicts ~2300 MWh (gross) annual production

• Given 10346 m2, 100% plant availability

– Similar to SolarGenix prediction • Annual output can be improved through seasonal changes in HTF flowrate

– Base case 42300 kg/hr all year – Low HTF temps in winter delays startup of ORC (need ~190o C)

and reduces turbine kW output after startup – It appears at least 2 separate flowrates are called for

30 32 34 36 38 40 42

January 2nd

hr

HTF Temp DNI

1000– Operate Nov - Feb at reduced flowrate 900

700– Rest of the year at full flowrate 800

600 500– Optimization will require further study 400 300 200 100

0

4

Future Saguaro(constant temperature, with storage)

Future Saguaro (constant temperature, with storage)

STEC storage control algorithm determines storage/ORC flow split and mixed return temperature to solar field

New Type 502 thermocline storage created from TRNSYS Type 10. Validated with Solar One data. Uses 23 differential equations.

Existing STEC Type 197

Derived from ASPEN model (NREL):

Out_Temp = f(In_temp, flow).

Dispatch stored energy during APS peak

Out_kWe = f(In_temp, flow),

5

Thermocline StorageBackgroundBackground

•

Thermocline Storage

Studies indicate that a thermocline-type energy storage should be the most cost effective option for parabolic trough power plants

• Thermocline was demoed at Coolidge ORC plant, 1979 to 82

– Stratified oil with no rock – 200 m3, H/D ratio = 3.5 – 68 % charge/discharge daily thermal η

• Thermocline was demoed at Solar One, 1982 to 86 (fire)

– 78% rock/sand, 22% oil on volume basis – 3460 m3, H/D ratio = 0.75 – 97% daily, 92% annual η (tank only)

• Nexant proposes a scaled-down Solar One tank for Saguaro

– 330 m3, H/D = 1.2 • All 3 have similar max operating temperature (~300 oC) 6

Solar One Thermal Storage

7

Solar One Thermal Storage

Fiberglass Insulation -Wall, 12 inches -Roof, 24 inches

Active Height modeled in TRNSYS

}Floor thermal mass included in new TRNSYS model

Validation of Type 502 thermoclinemodel with Solar One cooldown data

recorded November 5 to 22, 1982

150.0

170.0

190.0

210.0

230.0

250.0

270.0

290.0

310.0

330.0

0.00 2.00 4.00 6.00 8.00 10.00 12.00

StartSolar 1TRNSYS

Validation of Type 502 thermocline model with Solar One cooldown data

recorded November 5 to 22, 1982 Te

mpe

r atu

re (C

o )Te

mpe

r atu

re(C

o )

16.516.5 DayDay CoCooldown of Solar One Thermoldown of Solar One Thermoocline Stocline Storrageage

330.0

310.0

290.0

270.0

250.0

230.0

210.0

190.0

170.0

150.0

Start Solar 1 TRNSYS

0.00 2.00 4.00 6.00 8.00 10.00 12.00

Distance from tank bottom (m)Distance from tank bottom (m)

8

Validation of Type 502 thermoclinemodel with Solar One discharge data

recorded June 28,1983

150170190210230250270290

0.00 2.00 4.00 6.00 8.00 10.00 12.00

Sol 1 Midnight T (deg C)Sol 1 at 4 amTRNSYS at 4 amSol 1 at 8 amTRNSYS at 8am

150170190210230250270290

0.00 2.00 4.00 6.00 8.00 10.00 12.00

Sol 1 Midnight T (deg C)Sol 1 at 4 amTRNSYS at 4 amSol 1 at 8 amTRNSYS at 8am

Validation of Type 502 thermocline model with Solar One discharge data

recorded June 28,1983

150 170 190 210 230 250 270 290

Sol 1 Midnight T (deg C) Sol 1 at 4 am TRNSYS at 4 am Sol 1 at 8 am TRNSYS at 8am

If Solar One flowmeter was 25% too high

Tem

pera

ture

(Co )

Tem

pera

t ure

(Co )

0.00 2.00 4.00 6.00 8.00 10.00 12.00

290

270

250

230

210

190

170

150

Distance from tank bottom (m)

0.00 2.00 4.00 6.00 8.00 10.00 12.00

Sol 1 Midnight T (deg C) Sol 1 at 4 am TRNSYS at 4 am Sol 1 at 8 am TRNSYS at 8am

Distance from tank bottom (m)

If Solar One flowmeter was 15% too high

9

Discharge Test Validation Discharge Test Validation

• SAND86-8175 states that Solar One flowmeters were apparently reading 15 to 25% high during charge/discharge tests

• TRNSYS validation also suggests that readings were too high by ~20% and validation is based on this assumption

10

Final Solar One validation parameters. Physical properties defined most parameters. Three

parameters were “dialed” to match experiment.

Final Solar One validation parameters. Physical properties defined most parameters. Three

Parameter Solar One Value (Saguaro)

Units

0.669

Length of rock bed 38.6 (23) ft

2826 (415) ft2

188 (72) ft

0.244

165 lbm/ft3

0.041 Btu/hr-ft2-F

Axial thermal conductivity 1.28 Btu/hr-ft

0.015 Btu/hr-ft2-F

Floor Loss coefficient 0.18 Btu/hr-ft2-F

Oil density 49.3 (45.9)

Void fraction 0.22

1.5

lbm/ft3

parameters were “dialed” to match experiment.

Specific heat oil Btu/lbm-F

Cross-sectional area

Perimeter

Specific heat rock Btu/lbm-F

Rock density

Wall Loss coefficient

Roof Loss coefficient

Floor capacitance multiplier

11

Initial Simulation Insights for Future Saguaro

Initial Simulation Insights for Future Saguaro

• TRNSYS predicts ~3900 MWhe (gross) annual

– Given 18800 m2 (expanded field), 100% plant availability – Given NEXANT base case storage dimensions and operating strategy

– Stop storage charging at 226 oC oil exit temperature – Stop storage discharging at 193 oC oil exit temperature

– Gross annual efficiency of plant with storage (8.3%) somewhat lower than plant without storage (8.9%)

• Annual output of plant can be improved through some relatively minor optimizations

– Increasing storage volume by 50% increases output by ~250 MWhe – Active diameter and length are increased from 23 to 26.3 feet

– Relaxing dispatch strategy and increasing maximum charging temperature to >250 oC, may add ~75 MWhe

– Given these improvements, annual efficiency restored to ~8.9% 12

•Conclusions

The thermocline storage system proposed for Saguaro should perform well

– Install it in combination with the proposed field expansion – Capacity factor of plant will increase from 26% to 48% (given 100%

equipment availability) – However, additional control complexity may make “unattended operation”

impractical

Future Work • Validate TRNSYS model with Saguaro performance data

• Expand/improve models

– Include thermal losses of non-solar field piping – Include the effect of wet-bulb temperature on ORC performance – Plant parasitics

• Performance optimization studies 13