Discounting An Existing Solar System in Detailed PV with Storage

Hi Paul and forum group,

I've read in few areas on the Forum that SAM does not yet allow for using the existing PV system with its production data along with the load of the building that has the PV and modeling the addition of storage.

A few solutions I have come up with to try and discount the existing system while modeling on addition PV and first-time storage. I was wondering if you saw any issues with these approaches, which one seems the most accurate and if they seemed within bounds of SAM's logic. I am asking because I coming up with very different results. Note: this is a zero energy building scenario so we are working in negative returns at the moment because of net metering rules where we are.

1. Estimate native load by removing the impact of existing PV. Model as a complete new system, with existing PV represented in the total new capacity. Add in a CBI value on year 0 that is equal to what the existing PV is worth. Concerns: the cash "incentive" is being overvalued somehow in a 25yr analysis.
2. Take actual load that has the impact of existing PV (with some 15 min values reach negative because of peak summer production), adding in additional PV, storage and running. Concerns: SAM seems some days with a negative grid load but doesn't see the potential to capture that energy with the battery. It only sees the impact of new PV.
3. Model storage only (add in a tiny PV system as place holder--a few kW) with actual building load that has the impact of PV left in. Take results and combine with a PV only run of additional solar. Concerns: same issue as above.

Hopefully this makes sense. Would be happy to share files if you would like. Thanks for your time,

Henry

Hi Paul,

Thanks for your note-as usual very helpful and timely. Thanks so much!

A few clarifying items from my question and how I will be interpreting your response.

The facility I am looking at already has PV on it. I created a load profile for the facility from utility and solar 15 min data, which I am calling "native load" because, as you noted, it removes any excess generation back to the grid giving a picture of the facility's actual energy use. I am using this "native load" to model the existing PV (78kW) with an additional 280kW along with storage, or in another scenario, just storage in various financial arrangements. I want to model all the PV not just new arrays so that the battery controller knows what solar resources are available at each timestep.

My main question, which you seemed to have gotten to, is how to discount the cost for the existing PV on the financial side for the various setups. My concern was that because there is diminishing marginal value of the PV as the system output outpaces what the facility needs (in hopes of reaching zero energy), that I might need to be careful where and how I discount the existing PV costs. 

For the scenario which only sees us adding a battery, I will do as you say and just set capital costs of a PV system that match what is already on site to zero. For scenarios where I add PV, I will discount capital costs (rate per installed watt) that represent what is already onsite.

Thanks again.

Best,

Henry

Hi Paul,

Coming back to an older question and post. I am still wrestling with how best to model financials of adding storage to existing PV.

If I run as suggested above and set the direct capital cost of PV to 0 (sized to represent what is already existing at the site), isn't the model assuming that you're getting all the kWh from the PV with no cost, and thus skewing NPV's and cashflow? If SAM is looking at cost with/without system, then wouldn't it always be skewed towards storage? 


Thanks for your time.