Loading…
Net load is the load plan with wind and solar taken out. It is what dispatchable plant actually has to cover, and it is the curve that decides which hours are tight.
Two series are published while the day-ahead market is still open: the load plan and the wind forecast. EPİAŞ publishes no solar forecast, so solar is shaped from the mean realised profile of the seven preceding days. Bands are quintiles of the last thirty days rather than of the target day, so a flat day reads as flat instead of being rescaled to look dramatic.
Hour by hour
Why this curve and not demand
Demand peaks in the early afternoon in Türkiye through the summer. Net load usually does not: solar is at its strongest in exactly those hours, so the system is at its most comfortable when the meter reads highest. The tight hours arrive after sunset, when solar leaves in about an hour and demand is still near its plateau.
That gap is the whole scheduling problem. On a typical August day the load plan peaks around 14:00 while net load peaks around 20:00, and between those two points the system has to climb roughly 2,400 MWh every hour for eight hours. Reading demand alone hides that climb entirely.
Method
- NET LOAD
- Load plan minus wind forecast minus solar.
- SOLAR
- No EPİAŞ forecast exists. Shaped from the mean realised profile of the seven preceding days, which includes distributed rooftop output derived as consumption minus metered generation.
- BANDS
- Quintiles of the last thirty days, not of the target day.
- SCOPE
- Descriptive analysis of published data. Not advice, not a forecast, not a trading recommendation.
Actual vs outlook
Every net load outlook is published before the day-ahead price exists. The next day the same hours can be checked against what actually cleared, which is the only honest way to know whether the method is worth anything.
The outlook never forecasts price. What it claims is structural: which hours sit at the bottom of the range and which sit at the top. Those claims either hold or they do not, and the record below is kept whether they do or not.
What a failed check means
The method has three ways to be wrong, and each points somewhere different. If the low block lands in the right hours but the price does not follow, the bid stack is doing the work rather than the fundamentals. If the block itself lands in the wrong hours, the solar shaping is off, which is the weakest input because no forecast is published. If the evening block is right but the ramp is not, the wind forecast missed.
Recording which of the three fails is more useful than recording whether the day was right.
Loading…
Published after settlement data lands, from EPİAŞ open data. Every number is read from tool output rather than typed from memory, and the balancing figures separate market imbalance from grid constraints.
Generation mix
The hours that cost most to balance
| Hour | Day-ahead | Real time | Volume | Cost |
|---|
Against recent days
| Wind | Average | Floor | Spread |
|---|
Loading…
Türkiye's electricity market on one page: what happened yesterday, what we expect tomorrow, and how often our expectations turned out right. Point at any chart (or tap it on a phone) to see the numbers behind it.
Wholesale electricity prices (TL per MWh) set the day before by the day-ahead market, and what actually happened on the grid.
Where the electricity came from
Share of the day's electricity by source. Renewables are hydro, wind, solar and other clean sources.
Last-minute corrections
Plans never match reality exactly, so the grid operator orders plants to produce more (up) or less (down) in real time. More corrections means the day was harder to predict. The table shows the three hours where corrections cost the most: DA is the day-ahead price, RT the real-time price.
Compared with recent days
How yesterday differs from the day before and from the same weekday a week earlier. ▲ higher, ▼ lower; the bar shows how big the change is.
| vs day before | vs week before |
|---|
Published before tomorrow's prices are set. Net load is demand minus wind and solar: the part coal, gas and hydro must cover. Hours with the least net load are usually the cheapest, hours with the most are usually the dearest.
Net load outlook
Every outlook makes four claims that can be checked once prices are known: (1) the cheapest hour falls in the blue band, (2) the cheapest four hours in a row start there, (3) the dearest hour falls in the red band, (4) the dearest four hours in a row start there. The day after, each claim is marked right or wrong. Days without a red band carry only the first two claims.
Pick the days to look at: drag the blue handles or press 7D, 14D, 30D. Grey bars are each day's four-hour spread. It changes every panel inside this blue frame, each marked with the same date tag. The sections above the frame always show yesterday and tomorrow.
Four-hour price spread, day by day
For each day: the average price of its four most expensive hours in a row, minus the average of its four cheapest. Roughly what a battery earns per MWh by charging in the cheap hours and selling in the dear ones. Taller bar = bigger gap between cheap and expensive hours. The white line is the seven-day average.
How often the outlook was right
Of all claims made in the past seven days, the share that turned out right. 75% means three in four. The dashed line is the average over the days you selected. Rising line = the outlook is getting better.
Day by day
One square per day, oldest first. Colour shows how many of that day's claims were right.
Which claim is most reliable
For each of the four claims, how often it was right over the selected days.
Is the outlook better on some weekdays?
Checked days as a table
| Day | Low block | Top block | Held | 4 h spread |
|---|
Published analysis
Every post, indexed by the day it analyses.
The net load outlook covers the day ahead, so its analysis date runs ahead of its publication date. Market analysis covers a closed day.
Loading…
Energy news
The energy headlines from Türkiye that can move the power price, with what each one means.
Only news that has a path to the electricity price makes the cut. Each item says which way it pushes the price and why. Daily shows only today's post; the page turns over at 00:00 UTC.
Loading…
Last seven post days
Every Türkiye post from the last seven post days: the day-end scorecard, the net load outlook and the energy news.
Today's posts are under Daily. This page holds the seven days before today that had at least one post. Days without a post are skipped, so the list can reach back further than a calendar week. Open a post to see its chart, its clip and the full text.
Loading…
Market analysis
What happened in the British power market on the closed day: the price, and which sources kept the lights on.
The system price is what a generator or supplier pays or receives for every MWh it is out of balance. Daily shows only today's post; the page turns over at 00:00 UTC.
Loading…
Day-ahead outlook
How much wind Great Britain expects tomorrow, hour by hour, before the day starts.
Wind is the source that swings most in Great Britain: a windy hour usually means cheaper power, a still hour dearer. This is the system operator's own forecast, not a promise. Daily shows only today's post; the page turns over at 00:00 UTC.
Loading…
Last seven post days
Every Great Britain post from the last seven post days: the closed-day analysis and the day-ahead wind outlook.
Today's posts are under Daily. This page holds the seven days before today that had at least one post. Days without a post are skipped, so the list can reach back further than a calendar week. Open a post to see its chart, its clip and the full text.
Loading…
Two markets, four differences
GB and Türkiye run the same sequence — day-ahead, intraday, balancing — but the mechanics diverge in ways that change what an imbalance actually costs.
Written for people who know one of these markets well and need the other quickly. Times are from the exchanges' own published schedules.
Where the day is decided
GB closes its day-ahead auctions almost three hours earlier than Türkiye, which means a British generator commits to a position while its weather forecast is that much less mature.
| Time | Türkiye | Great Britain |
|---|---|---|
| 09:20 | — | EPEX hourly day-ahead closes |
| 09:50 | — | N2EX hourly day-ahead closes |
| 12:30 | GÖP bidding closes | — |
| 14:00 | Final MCP published | — |
| 14:00–15:30 | KGÜP generation schedules notified | — |
| 15:30 | — | EPEX half-hourly day-ahead |
| 17:30 | — | Intraday auction 1, 48 contracts |
| 08:00 (D) | — | Intraday auction 2, 30 contracts |
| Delivery −60 min | GİP closes | Balancing mechanism gate closure |
Four structural differences
| Türkiye | Great Britain | |
|---|---|---|
| Settlement period | 1 hour, 24 per day | 30 minutes, 48 per day |
| Day-ahead venues | One exchange, one price | Two exchanges, two prices |
| Imbalance pricing | Dual price: min and max of MCP and SMP | Single price, so direction decides gain or loss |
| Price cap | 4,500 TL/MWh ceiling | No equivalent ceiling; negative prices occur |
GB did not move to 15-minute market time units in October 2025. That change covered eleven coupled markets and excluded Great Britain and Switzerland, so the half-hourly settlement period stays.
Why the last two rows matter most
Under dual pricing every deviation is a loss, so the objective is simply to be accurate. Under a single price, being wrong in the direction the system needs is paid, so the objective changes: a portfolio can be deliberately short into a short system and earn from it. The same forecast error produces a penalty in one market and a payment in the other.
The price cap works the same way in reverse. A ceiling means scarcity stops being priced past a point, so the day-ahead market can clear at the cap for hours while real time settles far lower, which is exactly what happened in Türkiye on 26 August. Without a ceiling, GB lets the price go where the last accepted offer sits, and lets it go negative when there is too much.
Why direction matters more than accuracy
GB settles imbalance at a single price. A generator that comes in short when the system is short pays that price; a generator that comes in short when the system is long is paid. The same 2% deviation can therefore cost several times more for one unit than another, and the ranking by deviation size is not the ranking by cost.
Türkiye settles at a dual price, where a deviation is valued at whichever of the day-ahead and balancing prices is worse for the generator. Direction carries no reward there, only the size of the error matters. That is the single largest difference between managing a portfolio on each side, and it is why the GB scorecard reports a column that has no Turkish equivalent.