User Defined Power Block- Molten Salt Tower

Dear Sir,

I defined a power block performance look-up (user-defined) table in SAM V2017.9.5. I got the following error when simulating single owner molten salt tower simulation. I am not sure what is the reason of this error from the error message. Can you guide me debugging this error?

exec fail(tcsmolten_salt): At time = 2129, CR_DF__PC_MAX__TES_OFF__AUX_OFF failed to find a solution to achieve a PC thermal power less than the maximum
Simulation tcsmolten_salt failed.
SSC requires input 'gen', but was not found in the SAM UI or from previous simulations
Simulation singleowner failed.

I see sometimes unrealistic time as follows, 

exec fail(tcsmolten_salt): At time = 612000, CR_OFF__PC_TARGET__TES_DC__AUX_OFF failed
Simulation tcsmolten_salt failed.
SSC requires input 'gen', but was not found in the SAM UI or from previous simulations
Simulation singleowner failed.

Thank you. 

Hi Dhinesh,

The current version of SAM is SAM 2018.11.11. Please use that version to be sure you are using the latest models.

It is likely that some input parameter is incorrectly specifed. Please see the draft description of the model on the CSP Publications page under "CSP Power Cycle Models:"

sam.nrel.gov/concentrating-solar-power/csp-publications.html

Best regards,
Paul.

Dear Sir,
Thank you for your response. I switched to SAM 2018.11.11. But, I noticed that the new version appears to be unstable with the user-defined
power block. SAM more often closes automatically without any message/error. I checked this in two different machines and observed that SAM crashed in both of them randomly.

Moreover, SAM also reports error similar to the below one although the power block parametric table was developed from the SAM generate
table command and the table has 580oC row as the range was not changed.

exec fail(tcsmolten_salt): The user defined power cycle tablecontaining parametric runs on the hot HTF temperature must contain the upper
level HTF temperature 580 [C]. The maximum value is 574 [C]
Simulation tcsmolten_salt failed.
SSC requires input 'gen', but was not found in the SAM UI or fromprevious simulations
Simulation singleowner failed.

I am able to reproduce the above error by using the default user-defined model as follows.
1)  Copy the default user-defined power block loo-up table externally
2) reduce the number of rows to 10 for each parameter (HTFtemperature, mass flow and ambient temperature) in the user-defined power block
by maintaining the same range. This reduces the table size from 180 rows to 90.
3) copy the parametric table rows and generate the net_powerand heat input factor by interpolation from the default user-defined model.
4) run SAM with the reduced power block look-up table size and default solar filed settings

By doing this, I often faced the above reported error and also I noticed that SAM crashes without any warning.
Please advise me on this.
Thank you.
 

Hi Dhinesh,

Would you mind sharing a .sam file with your power cycle data so I can investigate further?

Best regards,
Paul.

Hi paul,

 I encountered similar problems when I tried to use a user-defined power block for the molten salt power plant. The swich between different versions of SAM doesn't have any effects on the results. In fact, several sets of data for the user-defined power plant work well. But for the cases with other sets of data, the simulations always break off due to the error "exec fail(tcsmolten_salt): At time = 1.39896e+07, CR_OFF__PC_TARGET__TES_DC__AUX_OFF failed Simulation tcsmolten_salt failed." 

Attached are two cases with different sets of data for the user-defined power plant, the one named "Flex_no_opt_32" works well while the other one named  "Flex_no_opt_41" always encounters errors. 

I still have several sets of data for thefollowing  user-defined cycle simulations. So I will appreciate if you can explain how this user-defined cycle works during simulation, like how is it integrated in the simulation process.

Besides, The look-up table for user-defined cycle has lots of limiit espacially for the cycle with unique performance like the one in my case. Is it possible to defined a user defined cycle in other platform like Matlab and connect it with SAM to use the annual-simulation function. I know SAM is now an open-source software. Is there any examples of cross-platform simulation cases, especially matlab-SAM?

These questions are of much importance to my Ph.D work, your help will be greatly appreciated.
Looking forward to your reply.

Best regards,
Juergen Ma

Hi Dhinesh,
 
Thank you for your detailed description of the error.
 
The UDPC model in SAM 2018.11.18 contains some bugs related to how it sorts the data into the 9 parametric sets shown in the help. There are some rules to follow for this release to avoid the bugs and the resulting hard crashes and table constraint errors:

• Each parametric should contain at least 4 rows
• Each parametric should NOT contain the design point value of the parametric variable
• Each parametric SHOULD contain the low and high level values of the parametric variable, and they should be the lowest and highest values, respectively, in that parametric.

 
These bugs are fixed in the upcoming SAM release, and in that release the UDPC model will not require the 2^nd or 3^rd rule listed here. It will still require that each parametric contain at least 4 rows.
 
Ty