Difference between Lifetime and Annual (first year) outputs in PV+Battery

Hello SAM Team!

Despite knowing that SAM does not (yet) model off-grid systems, I consulted on the forum some tips that were answered to try to model such systems using the current version (links:  sam.nrel.gov/forum/forum-general/3182-photovoltaic-plant-storage-best-option-for-dispatch-model.html ,  sam.nrel.gov/forum/forum-general/2998 )

In "Grid limits" I set to "Enable interconnection limits" with 0 kWac and I entered a curtailment array with 0.0000001 MW at all hours of the year, in an attempt to reset the interactions with the network.

In addition, in "Electricity rates" I set "buy all / sell all" with an entry with a zero sale rate.

I am having trouble interpreting the results, since the values ​​of import and export of energy from the grid / to the grid (considering, respectively, how the power of the diesel generators drive and wasted energy) are not the same in the hourly simulations ( for the first year) and lifetime (considering the first 8760 hours).

For example, data from the first year account for "Electricity to grid" null but in Lifetime data, it is 14532 kWh for the first year.

I'm sending my file . I am using the generic system model.

Can you please help me understand the results? What would the system energy be? Photovoltaic generation (in my case, that I added the matrix in "Power plant") and the battery? What energy flow does it account for?

In addition, does the year that the SAM simulates consider the 2020 working days and weekends calendar? Or rather, what day of the week is the first day of the year for the simulations?

Thank you!

Sincerely,

Luiz Pereira

Hi Luiz,

As you note, SAM assumes that the system is connected to the grid, and it is not always possible to prevent the model from using the grid.

The "lifetime hourly data" outputs are the outputs of the performance model, in this case for the Generic - Battery system. The "hourly data" outputs are outputs of the financial model, in this case the Commercial model.

The hourly "electricity from system to load (year 1 hourly) (kWh)" output is equivalent to the sum of the lifetime "electricity to load from battery (kW)" and "electricity to load from system (kW)".

You can use the "system power before grid curtailment (kW)", "system power before grid interconnect (kW)", and "system power generated (kW)" to explore the effect of the interconnection and curtailment inputs.

SAM assumes that the first day of the year is a Monday, and that weekends are Saturday and Sunday, regardless of the year time stamp in the weather file. It does not account for daylight savings time or leap years.

Best regards,
Paul.

Hi Paul!

Can you help me in another issue? 

In DC-connected systems (Battery Storage > Detailed PV-Battery > Distributed > Commercial Owner), what does means the "system power generated" output in Lifetime Hourly Data? When I compare the profile curve there's "generation" when there is no sunlight, what make me understand that this output is like a "black-box" which accounts both the PV and battery energy. Is there any output that provide just the PV system energy or not? How the energy is accounted in this output (how energy flows)?

Thank you!

Hi Luiz,

As you note, the "system generated output" variable includes power from the battery, so is a bit confusing. We are planning to add some additional variables to make this clearer: github.com/NREL/SAM/issues/370 .

Best regards,
Paul.

Ok Paul, thank you for considering that for a next release.

In my research, I'm using the DC-connected simulations to measure how the PV+Battery system allows a building to become self-sufficient. For this, I'm using the Self-sufficient rate (the sum of PV generated and exported for the grid over the sum of energy consumption in an specific time) and Self-consumption rate (the sum of PV generated and exported for the grid over the sum of energy generated in an specific time). I'm realizing that considering periods of time of years or months, I'll not have problems in the results (the sum of system generated probably will not change even so confusing), but reducing the periods of time for days or hours, do you think my result would be trustworthy or the hourly confusing data can deceive the results and it is better to lay down this results?

Let me know what you think, please.

Thank you!

Hi Luiz,

I think you should be able to size the system and design a dispatch approach that minimizes exports and imports from the grid on an hourly basis.

Best regards,
Paul.