Modelling fuel savings, controlled (curtailed) PV on a diesel grid
I would like to model the economic cost/benefit (rather than system stability) of adding controlled PV to a diesel grid. By controlled PV, I mean a PV plant which can be curtailed by the network operator if required. Reasons for curtailing PV output might include maintaining generator minimum loading and spinning reserve sufficient to handle sudden cloud passages.
Can HOMER be used to model the expected fuel savings in this scenario? I note that within HOMER generator minimum load ratios and fuel curves can be specified, and spinning reserve can be constrained as a proportion of PV output for the current timestep. However controlling PV might help meet this spinning reserve constraint.
For a made-up example, consider a 5MW load, online is a diesel generator which must operate between 4MW and 6MW, and a PV plant with a current output of 2MW. Let's say that spinning reserve must be sufficient for a 10% change in load and total loss of PV output. Will HOMER bring another generator online (with minimum load >1MW), and hence not use any PV output? Whereas if the PV plant is curtailed to 1MW at this time, the 1MW PV output can be made use of. Can I model this?
Any advice as to whether HOMER is the appropriate tool for calculating fuel savings in this sort of scenario would be much appreciated.
HOMER can do fuel savings calculations, similar to how you envision.
It won't be exactly as you envision, however, because HOMER doesn't curtail PV explicitly. Any extra energy produced by the PV is counted as "Excess Electricity". So if you model your example system, you'll get lots of the PV energy being counted as 'Excess', meaning it is not serving the load, and this 'Excess' can be thought of as being curtailed.
Your other results will be correct: the LCOE is not affected by how much energy goes to excess. The overall system LCOE only considers load served, so your LCOE won't be skewed or appear cheaper, just because lots of your energy it 'excess'.
Bottom line: HOMER is the appropriate tool. Hope that helps!
Hi Steffi, thanks for the reply. That makes sense that HOMER can spill solar to maintain minimum generator loading. However I am still a bit unsure how to deal with ensuring the correct amount of spinning reserve is maintained during HOMER's generator dispatch decisions.
I'll take the numbers I provided above, and for now conservatively assume PV output could drop totally and instantly with cloud passage. If I specify within the constraints tab of HOMER that operating reserve must include 100% of solar output and 10% of load, then am I right in thinking that HOMER will decide operating reserve needs to be 2MW (current PV array output including excess electricity) + 0.5MW (10% of load) therefore bring another generator online (because 2.5>2) and spill all solar due to minimum loading of 2 generators? Which is not what I want, I want to save fuel because I can curtail PV output to say 1 MW and only require operating reserve of 1MW (utilised PV output) + 0.5MW, which I have available with the one generator.
If I set operating reserve includes 0% of solar output and 10% of the load then that scenario works fine. However now consider the same scenario except the load is 5.6MW. HOMER sees no problem with the load being met with 4MW from the generator and 1.6MW from the PV output. However I am now not maintaining sufficient spinning reserve to cover loss of the 1.6MW PV output + 0.56MW load variation.
Am I understanding this correctly, and is there a way to model this?
HOMER works like you expect. By specifying operating reserve of 10% of the current time step's load and 100% of the solar power output, the model must meet the load+10% of the load + 100% of the solar output. Assuming the your 6MW generator and 2MW PV cannot meet the load of 5MW+2MW+0.5MW, it will bring on another generator, if you have allowed the possibility or 2 generators. Then it will 'spill' all the excess electricity coming from the generator or PV into the 'Excess Energy'.
See this screenshot of how HOMER works? I made the system you described with the constraints you described. Seems to work without allowing a second generator. The graph shows the load, the required operating capacity, and exactly how the generator is controlled.
Obviously, the best way to understand how HOMER works is to try it out yourself.