Open simonhac opened 7 months ago
I'm doing some stats homework using the NEM data for batteries this week asking these kind of questions. I'd appreciate any improvements, even though it wont be in time for me. I'll have some graphs to share in two weeks time.
beyond cycles of charge and discharge or 'turns' I'm interested in how many times a battery is discharged to it's operational bottom, or within 10% of it's operational bottom and how long before it's fully charged, or within 10% of fully charged. And what kind of RE resources are available in those gaps. For the most part just scanning the data, it looks like with so much price setting by fossil gas that the sky's the limit for new deployments of batteries, they're barely making a dint on gas (OCGT, ICE, steam, CCGT) exports even in SA at the moment.
I realise that they're earning a lot of income from ESS/FCAS, but even just considering their wholesale buy/sell spreads, their cost to purchase the power is ~10% of their income from their wholesale selling of power.
I aggregated ~14 months of 5 min interval NEM wholesale trade data into a costs and income statement (ignoring everything other than energy trades) by month for each NEM state, see below. buy cost seems be ~10% to ~900% mark up is a pretty handy margin in most industries. And if gas stops being a price setter, batteries will become a price setter (could be an issue one day, a nice problem to have I guess).
I'd like to know if the energy that batteries use for ESS/FCAS service bidding also has to be bought or sold on the wholesale market to cover the energy imported (Frequency down) or exported (ƒ up)? Can anybody explain that part of the market to me or refer me to some easy to understand info somewhere?
should OpenNEM be exposing ESS/FCAS market data?
I looked on AEMO and found the relevant data files but it's one CSV file per (5 min? I forget the interval of FCAS files) and a bit tricky to understand the data inside the spreadsheets when compare to wholesale market files.
Still need to eyeball a bunch of rows to validate my SUMPRODUCT formula to check validity but interesting so far.
Needs design work, but some early thoughts:
1. The
battery storage capacity
in MWh should be clearly displayed.2. Users are interested in the arbitrage value and the income of batteries.
discharging income
lesscharging cost
. in many cases the charging cost will be negative, so here themarket value
is actually $13.5k + $3.7k = $17.2k.average net income
— the spread? this would be[discharging income minus charging cost] / discharging volume
.3. Users are also interested in the number of "turns" of the battery.
In this case, the battery discharged 37,698MWh in the quarter, which for a 450MWh battery, is 83.8 turns, or 0.91 turns/day.
note that over shorter periods these numbers become nonsense — since the amounts of the energy in the battery at the start and end of the period are unknown.