Hi Yasir,

I think you need to be more specific in your question. Are you looking for a method to determine the optimal capacity of the battery? Is the power rating already fixed by the inverter size? Is there a cost element to this question?

Best,

Diarmid

Diarmid Roberts Thank you for your kind response. I am actually working on a Simulink model for a BESS controller. It is implemented in a grid with 4 x 900MVA Synchronous generator units. The power rating isn't fixed but the inverter rating is 1100VDC/690VAC (using an Average model of Inverter). There is no cost element involved.

The primary purpose of the BESS is to provide Inertial and Primary frequency response in case of frequency disturbances amid Generation-Demand imbalance or Short circuit faults. Thus, I am looking for a sizing methodology for BESS in this regard. I have been through some research papers but they mostly consider algorithms that involve cost minimization problems.

Yasir,

Are Inertial and Primary frequency response separate functions/services? Is one more important than the other?

Diarmid

Diarmid,

Yes, both of these would be separate services with priority given to Inertial response.

Regards

I would suggest just running your simulation at different MWh ratings to find the minimum that satisfies your resilience target across a sufficiently large set of scenarios. The duration required for the BESS will depend on the distribution of frequency deviations - I know that in the UK the deviations in frequency can have multihour periodicity, so you probably want to look in the range 1 to 4 hours to start with.

Hope this helps,

Diarmid

Thank you very much for the comprehensive answer. I really appreciate that.

Just one additional factor that I wanted to consider along with this was battery SoC. Normally the BESS should operate between 20-80% SoC so just to make sure I understand your point correctly, keeping that SoC range in mind, should I consider running simulations in such a way that for example if the initial SoC is 80% then by the end of this service period (say 4 hours) it should be somewhere around 20% and not below that?

As per my understanding, Inertial response duration is of the order of seconds while Primary response may last up to 15 mins. But from what you have highlighted, it is really interesting to know that I should not only consider a single event rather an overall periodicity for the proper BESS response. I would be really grateful if you can kindly provide insight from this point of view.

Regards,

Yasir,

Yes, you need to consider SOC constraints, as that's a crucial aspect of the sizing optimisation - does the battery have enough headroom/footroom to perform the entire service period without emptying or filling? You should look closely at the requirements of the services in your jurisdiction. I know that in the UK, for Enhanced Frequency Response (the most dynamic service) the output of the battery is dictated by the deviation in frequency, and there is a 'dead band' when the frequency is close to the ideal, in which you can do what you want with the battery, e.g. charge it to recover SOC. There's a diagram of this in my colleague's paper: https://www.sciencedirect.com/science/article/abs/pii/S030626191831866X

Looking at longer time periods might not be as important to you if you are focussing on the control system, but it's important for higher level modelling as you don't want to get caught out unable to respond to a primary response call when you have already run down the SOC doing inertial response.

I should also clarify that I meant look at a range of battery capacities from 1 to 4 h (i.e. 1MWh to 4MWh in your case) - the time period you study could be even longer.

Thank you very much. That clarifies alot of things. I would go through the paper for the SoC constraints and relevant operating modes.

Regards