Gross cycle power output and hour of day mismatch for LFR

Hello Paul

Could you comment on the following mismatch between time of day and gross cycle power output in the attached SAM results file. The input file is also attached. The input is a constant DNI = 767 W/m2. This constant DNI helps isolate the mismatch effect from any possible vagaries of the usual DNI daily cycle.

The table below shows the first day from the results file. The gross output builds up to over 35 MWh before 12 noon then drops to just over 11 MWh at 1 pm.

The first issue is the lack of symmetry about 12 noon.
The second issue is that the LFR is a dynamic system. Therefore, I would expect the gross output to decline slowly into the afternoon.

Hour Gross cycle power output (MWh), hourly
1 0
2 0
3 0
4 0
5 0
6 0
7 0
8 2.12896
9 23.1518
10 32.8038
11 35.0451
12 35.2807
13 11.3254
14 8.00973
15 7.24113
16 6.25559
17 0
18 0
19 0
20 0
21 0
22 0
23 0
24 0


Regards


William Bell

Dear William,

It is hard to tell what might be causing the effect you observe without seeing the file. You can plot or view efficiency values along with the DNI and electrical output values to try to pinpoint the cause. Those variables are listed as "Collector optical efficiency," "Cycle conversion efficiency," and "Solar field thermal efficiency." One possible cause might be parasitic consumption.

The output of the default linear Fresnel case (and for the default parabolic trough case with no storage) tends to follow the DNI rather than lag behind it. Have you seen different daily generation profiles?

Best regards,
Paul.

Hello Paul

Thank you for your response.

---

It is hard to tell what might be causing the effect you observe without seeing the file.
The SAM input and output files are attached to the original message. Is there another file that I should attach?

---

One possible cause might be parasitic consumption.
The parasitic load in the table below is calculated from gross power less net power shown in the output file attached to the original message.
The parasitic load looks small and nothing out of the ordinary.

Hour Gross power (MWh) Net power (kWh) Parasitic (kW)
1 0 0 0
2 0 -166 166
3 0 -166 166
4 0 -166 166
5 0 -166 166
6 0 -166 166
7 0 -195 195
8 2 1,802 327
9 23 22,618 534
10 33 32,161 643
11 35 34,335 710
12 35 34,564 717
13 11 10,939 387
14 8 7,684 326
15 7 6,917 324
16 6 5,933 323
17 0 -211 211
18 0 -209 209
19 0 -169 169
20 0 -166 166
21 0 -166 166
22 0 -166 166

---

The output of the default linear Fresnel case (and for the default parabolic trough case with no storage) tends to follow the DNI rather than lag behind it. Have you seen different daily generation profiles?

Yes. I used Excel Pivot Tables to examine yield from the years 2007-12 to examine yield by hour. The clipping in the late afternoon is apparent when taking the maximum yield each hour over the month. However, this could simply be attributed to changes in DNI levels. Therefore, I isolated the effect using a constant DNI level in the SAM input file attached to the first email.

Additional information:
We use the default LRF setting for the Novatech solar boiler with the following four advised changes to the default.

Advised Default
Number of modules in boiler section: 17 12
Collector azimuth angle: -10° 0
Turbine inlet pressure: 120bar 90
Design turbine gross output: 30.07MWe 49.998

Regards

William Bell

Dear William,
It would be helpful for me to see a copy of the SAM file itself (.zsam) so that I can accurately replicate your analysis. It has all the input parameters for the linear Fresnel model. Also, the CSV "TMY Input" file you attached is not in one of the formats SAM can read and does not have all of the data I would need to convert it to the TMY3 format. For example, I would need the location's latitude, longitude, and time zone.
Here is the output from the Novatec solar boiler sample file for a few days in August that shows the shape you should expect from the model with reasonable inputs: 

Best regards,

Paul.

Hello Paul

Thank you for the extra information.

I have attached the zsam file to the original message.

The clipping of the expected block shaped yield could be related to SAM's calculations for the plant's non-zero azimuth orientation.

Regards

William

Dear William,

Thank you for sending the file -- it's very helpful for me to see all of the inputs.

On the Collector and Receiver page, it looks like you changed the "Optical characterization method" from "Collector incidence angle table" to "Solar position table" without changing the efficiency values in the table. That means that SAM interprets the angles in the table's x and y headers as sun position angles (sun zenith and altitude) instead of incidence angles (transverse and longitudinal incidence angles). That may explain why the output in the afternoon is not what you expect.

Best regards,
Paul.