Concerns regarding SAM's functionality and Stochastic Analysis

Dear Sirs,

Greetings.

I have some concerns regarding SAM’s Functionality, Solar Resource Library, and Stochastic Analysis. If possible, can you please answer my questions below?

1) I decided to simulate the weather conditions of the oldest CSP power plant (PS10 located in SPAIN). I checked the DNI value of the location using the plant's coordinates from the NSRDB Data Viewer. The actual DNI is different from the one SAM is calculating and using when importing the weather file to it. The actual value is 6.12 while SAM is using 3.91. I tried Default TMY file and Choose the year (here the year of 2020) options. Both are giving results with lower values. What is the reason behind this inconsistency?
2) Are the default locations in solar resource library according to TMY? If not, depending on what?
3) If I wrote the name of the country, SAM imports a weather file data. For example, when I entered USA, the weather file I got was usa_39.390897_-99.066067_psm3-tmy_60_tmy. So based on what the lat/long coordinates (here 39.390897 North, 99.066067 West) are selected?
4) In SAM’s Help, it’s written that the Single Owner LCOE calculation is different from the LCOE calculator. Can you please interpret that?
5) With regards to Stochastic Analysis, the mean and std. What are the reasons behind selecting normal distribution and those specific values? I converted those distributions into uniform functions. I did so because when I used the values in the literature, I got large negative values for the LCOE and the total annual power to grid. Is that correct? I got reasonable answers though.
6) When I run stochastic analysis, will SAM scan the Location and Resource and Heliostat Field pages? Because I want to run the stochastic for different locations and layouts.
7) How to neglect the effect of incentives, depreciation benefits, taxations, debt percent and other financial parameters? When I set them all to zero, I got an error for the LCOE simulations. In UAE, I don’t think there are such incentives. There is only 5 per cent on a taxable supply of goods or services. What I'm thinking of is that if I’m fixing all the variables other than the study variables for all the cases, I don't think the correlation between LCOE and the study variables will be different when removing the financial parameters. So please let me know your opinion here.
With regards to Power Purchase Agreement (PPA), I used the single owner model which is similar to the merchant model, except that the latter requires the spot market values to be defined. Other models concentrate more on equity, tax sharing and so on. So,
the system is very complex in terms of financing. I don’t want to include the complexity of these models when building my final model. What do you think?  
9) For Grid Interconnection Limit, as I understood, it’s the end-node electricity demand that controls the power production. Am I right?

Looking forward to hearing from you. Thank you.

Kind regards,
Nada

Hi Nada,

Please see your questions with my responses:

1) I decided to simulate the weather conditions of the oldest CSP power plant (PS10 located in SPAIN). I checked the DNI value of the location using the plant's coordinates from the NSRDB Data Viewer. The actual DNI is different from the one SAM is calculating and using when importing the weather file to it. The actual value is 6.12 while SAM is using 3.91. I tried Default TMY file and Choose the year (here the year of 2020) options. Both are giving results with lower values. What is the reason behind this inconsistency?

The NSRDB contains several different datasets and the DNI data in the different datasets is not the same. Also, are you comparing the same units? SAM displays an annual average value in kWh/m2/day on the Location and Resource page, which is different from the peak DNI value or the DNI at a particular time in W/m2.

2) Are the default locations in solar resource library according to TMY? If not, depending on what?

SAM's solar resource library is a collection of weather files on your computer that you add to the library.  Please see the "Folders and Libraries" topic in SAM's Help system. You can find it in the table of contents under "Weather Data."

3) If I wrote the name of the country, SAM imports a weather file data. For example, when I entered USA, the weather file I got was
usa_39.390897_-99.066067_psm3-tmy_60_tmy. So based on what the lat/long coordinates (here 39.390897 North, 99.066067 West) are selected?

You should use a precise location name such as a street address or place name, or the latitude and longitude (Lat,Lon) to identify a location. That latitude and longitude appears to be the geographic center of the United States. SAM uses a geolocation web service to translate place names to latitude/longitude pairs. You should get results similar to finding a location on Google maps.

4) In SAM’s Help, it’s written that the Single Owner LCOE calculation is different from the LCOE calculator. Can you please interpret that?

The single owner and other financial models calculate the LCOE based on an annual project cash flow as described in Help "Levelized Cost of Energy (LCOE)" in the "Financial Metrics" section. The LCOE calculate uses a fixed-charge rate method that is described in the "LCOE Calculator" topic under "Financial Parameters."

5) With regards to Stochastic Analysis, the mean and std. What are the reasons behind selecting normal distribution and those specific values? I converted those distributions into uniform functions. I did so because when I used the values in the literature, I got large negative values for the LCOE and the total annual power to grid. Is that correct? I got reasonable answers though.

SAM's Stochastic simulations inputs allow you to specify distributions appropriate for your analysis. Different distributions may be appropriate for different types of variables, and different kinds of analysis.

6) When I run stochastic analysis, will SAM scan the Location and Resource and Heliostat Field pages? Because I want to run the stochastic for different locations and layouts.

The Include weather file normal distribution based on DNI or DHI check box and DNI/DHI option make it possible to include weather data in the distribution. This requires a collection of weather files to represent the range of resource data. However, for your application, I think you would want to run a separate stochastic simulation for each weather file rather than use that option so that you get a separate set of results for each location.

For the Power Tower model, you should check Optimize heliostat layout and tower dimensions on the Heliostat Field page to ensure that SAM calculates a new field layout for each stochastic simulation. Note that this will make the simulation take a long time.

7) How to neglect the effect of incentives, depreciation benefits, taxations, debt percent and other financial parameters? When I set them all to zero, I got an error for the LCOE simulations. In UAE, I don’t think there are such incentives. There is only 5 per cent on a taxable supply of goods or services. What I'm thinking of is that if I’m fixing all the variables other than the study variables for all the cases, I don't think the correlation between LCOE and the study variables will be different when removing the financial parameters. So please let me know your opinion here. With regards to Power Purchase Agreement (PPA), I used the single owner model which is similar to the merchant model, except that the latter requires the spot market values to be defined. Other models concentrate more on equity, tax sharing and so on. So, the system is very complex in terms of financing. I don’t want to
include the complexity of these models when building my final model. What do you think?  

To simplify the financial model, you can use the PPA Single Owner model. To remove income taxes, tax credits, and depreciation from the analysis, set both the federal income tax rate and state income tax rate on the Financial Parameters input page to zero. To remove debt from the analysis, change the debt option to Debt Percent and set the debt percent to zero. You should also set the debt closing costs and up-front fee to zero if there is no debt.

9) For Grid Interconnection Limit, as I understood, it’s the end-node electricity demand that controls the power production. Am I right?

The grid interconnection limit sets the maximum value of "System power generated." If, in a given time step, the system generates more power than the limit, SAM sets the output power to the limit. The CSP system control algorithm does not take into account the grid limit.

Best regards,
Paul.

Dear Mr. Paul,

Thanks a lot for your reply.

I have other questions in my mind. If you can help me, I will be grateful.

1) Which is more reliable method to use, the Single Owner LCOE calculation or the LCOE calculator? Which of these two methods is used by stochastic analysis?

ANSWER: The PPA Single Owner financial model uses a relatively detailed pro forma cash flow to calculate metrics like net present value (NPV) and levelized cost of energy (LCOE). The LCOE Calculator financial model uses a simple fixed charge rate (FCR) method to calculate the LCOE. (This method is described in SAM's Help System, see "LCOE Calculator" under "Financial Parameters in the table of contents. (Press the F1 key or Help button to open Help.) Each method has its place, and one is not more reliable than the other.

2) How can I find the total field optical efficiency in SAM's results page? There are only the hourly data plus the field optical efficiencies versus Azimuth and Zenith angles.  Also, here, what are these angles referring to? Since the number of heliostats is high, I don't think they are reflecting the angles of the structures, am I right?

ANSWER: SAM does not report a total annual solar field optical efficiency for the CSP Power Tower model. The sun azimuth and sun zenith angles indicate the sun's position in the sky.

3) I want to study the effect of different heliostat patterns on two outputs, so I have to integrate SolarPILOT with SAM, in order to maximize the optical efficiency and lower the LCOE. The optimization tool in SAM doesn't consider parameters other than number of heliostats, tower height and receiver height and diameter. Also, I can't define the optimization function in both programs. Both are optimizing based on the LCOE. Moreover, both programs have common inputs as well as different parameters. SolarPILOT has more detailed input parameters for the receiver and the layout. Another thing to mention, there is no stochastic analysis tool in SolarPILOT and this makes it difficult to use parametric analysis for this huge amount of variables. I read about this CSP subject and learned these two programs in two months and decided to do my research on solar tower plants. I have ideas, but is there a gap I can't identify and fix? Can you please help me to sort this out? I don't know how to integrate them together because of the different parameters and because the optical efficiency and the LCOE will change when using them separately.

ANSWER: SAM runs SolarPilot when you choose the "Optimize heliostat layout and tower dimensions" option on the Heliostat Field page. As you note, the SolarPilot application itself has many inputs that are not available in SAM's user interface. You will have to determine how best to use the two models for your analysis given each of their capabilities and limitations.

Looking forward to hearing from you. Thank you again.

Kind regards,
Nada