Electricity marketforecasts

Forecasts of system demand, renewable generation and the day-ahead price (PTF). Model performance is measured daily against EPİAŞ Transparency data and reported.

MEGAWATTRON produces forecasts of system demand, renewable generation and the day-ahead market clearing price. Model performance is measured daily against the data published on EPİAŞ Transparency and reported.

View plans Request a demo

TR · published daily at 10:15 · actuals from EPİAŞ · UTC+3

7-day WAPE

pending

sample days

awaiting data

forecast · MWactual

Demandawaiting measurement
Wind – Licensedawaiting measurement
Wind – Unlicensedawaiting measurement
Solar – Licensedawaiting measurement
Solar – Unlicensedawaiting measurement
Hydroawaiting measurement
PTFawaiting measurement
Demandawaiting measurement
Wind – Licensedawaiting measurement
Wind – Unlicensedawaiting measurement
Solar – Licensedawaiting measurement
Solar – Unlicensedawaiting measurement
Hydroawaiting measurement
PTFawaiting measurement

Our areas of work.

Forecast products are offered by subscription; optimisation and software work is planned as a project around the organisation's needs.

Seven areas of work, starting with forecast products and extending to optimisation and custom software on the same data and model infrastructure. Forecast products are offered by subscription; optimisation and software work is planned as a project around the organisation's needs.

01Day-ahead and intraday market forecasts

Subscription

illustrative diagram

Demand
PTF
Wind
Solar
Hydro

Explore

Hourly forecasts of system demand, the day-ahead market clearing price, and wind, solar and hydro generation. Day-ahead forecasts are published every morning; intraday products are being prepared to update throughout the market sessions. Accuracy is measured daily against EPİAŞ Transparency data.
- Demand
- PTF
- Wind
- Solar
- Hydro
Explore
Subscription
Plant-specific hourly generation forecasts for solar, wind and hydro plants, to improve the accuracy of production schedule (KGÜP) submissions and to monitor imbalance cost. Actual generation is taken from EPİAŞ for licensed plants and from the organisation for unlicensed plants.
- Solar
- Wind
- Hydro
- KGÜP
Explore
Subscription
Hourly consumption forecasts at distribution-zone and substation level, or for a single site. Modelled on the organisation's meter and operational data to support supply planning, capacity management and lower imbalance cost.
- Zone
- Substation
- Site
Request a meeting
Project-based
A forecast-based decision layer for portfolios made up of generation, consumption and bilateral contracts: day-ahead and intraday bidding strategy, position management and control of imbalance risk.
- Bidding strategy
- Position
- Imbalance risk
Request a meeting
Project-based
Operational planning under grid constraints for local grids and organised industrial zones with distributed generation, storage and flexible loads; improving capacity utilisation and losses together.
- Distributed generation
- Flexible load
- Industrial zones
Request a meeting
Project-based
Charge–discharge schedules based on price and generation forecasts for storage facilities and co-located batteries. Arbitrage, capacity and ancillary service revenues are assessed together with battery lifetime constraints.
- Charge–discharge
- Arbitrage
- Ancillary services
Request a meeting
Project-based
Not an off-the-shelf product: monitoring, automation and decision-support software developed specifically for your organisation's plant, grid, site or trading processes. From data collection and SCADA/meter integration to reporting and automated workflows, it is designed and deployed to work alongside your existing systems.
- Monitoring
- Automation
- Integration
- Reporting
Request a meeting
Custom development

2 of 7 areas live by subscription · 5 areas project-basedTo discuss a project: info@megawattron.com

Forecast products

For seven products: an hourly forecast-versus-actual comparison of the last 3 days up to yesterday, with the measured error; for solar and wind the error is expressed relative to installed capacity.

Seven products for the Turkish electricity market: system demand, licensed and unlicensed solar generation, licensed and unlicensed wind generation, hydro generation and the day-ahead market clearing price. Each product comes with an hourly forecast-versus-actual comparison of the last 3 days up to yesterday and the error measured over that period; for solar and wind the error is expressed relative to installed capacity, for the others WAPE is used.

Daily error

Electricity Demand· MW

accuracy being measured

sample data

National electricity consumption forecast (hourly, MW)

Daily WAPE—

loading showcase dataPlant-level generation forecast

Accuracy measurement.

Daily error for the last 30 days, the period average and the best and worst day; measured against EPİAŞ actuals.

For each product we publish the daily error for the last 30 days, the period average, and the best and worst day. Measurement is made against EPİAŞ actuals; all results are reported, including the unfavourable days.

WAPE for demand, price and hydro: Weighted absolute percentage error: deviations in low-consumption hours do not overshadow the high-consumption hours. It is a widely used metric in energy markets; lower is better.
The measurement window is stated: Each value is shown with the number of hours and the date range it was measured over; a metric is never published on its own.
Installed-capacity error for solar and wind: Mean absolute error is expressed relative to the product's national installed capacity; every hour of the day is measured. Installed capacity is updated monthly from EPDK and EPİAŞ data.
Actuals come from EPİAŞ: The comparison is made against the official actuals of the EPİAŞ Transparency Platform.

Last 30 days · TR

Same calculation as the endpoint the dashboard reads · publication history began in August 2026

The daily forecasting process.

Data collection, modelling, publication, access and next-day verification; the same sequence runs for every product and plant.

A forecast day has five stages: data collection, modelling, publication, access and next-day verification. The process runs in the same order for every country, product and plant; the output of each stage is the input to the next.

00:0006:0012:0018:0000:00 +1

Early morning
Data collection
Actual demand and generation and KGÜP/UEVM submissions from the EPİAŞ Transparency Platform; hourly temperature, irradiance and wind forecasts from Open-Meteo; calendar and public holiday information.
input · EPİAŞ · Open-Meteo · calendar
Before publication
Modelling
Gradient boosting models trained separately for each product produce the 24 hours of the next day. The models are retrained regularly on current data and tested on historical data.
output · 24-hour forecast · confidence score
10:15
Publication
Next-day forecasts are published in the dashboard; the daily email report is sent at the same time. Plant-level forecasts are part of the same publication, as a separate series per plant.
dashboard · email · API
Throughout the day
Access
CSV export; automatic transfer to ETRM systems, Excel or in-house models with an API key. Web, mobile and integrations use the same data contract.
CSV · REST API · key-based access
Next day
Verification
As actuals arrive from EPİAŞ, the deviation of each hour and the WAPE of the day are calculated; the result is published in the dashboard and in the accuracy history.
hourly deviation · daily WAPE · 30-day history

Plant-level generation forecast.

A model specific to your organisation's solar, wind or hydro plant. For licensed plants, actual generation is taken from EPİAŞ; for unlicensed plants, the organisation uploads the generation history.

In addition to the national products, a model specific to your organisation's solar, wind or hydro plant. For licensed plants, actual generation is taken from EPİAŞ automatically; for unlicensed plants, the organisation uploads the generation history.

1
Application
A licensed plant is selected from the EPİAŞ catalogue; for an unlicensed plant, the location, installed capacity and generation history are submitted.
2
Approval
Our team reviews the application and the data; after commercial approval it moves on to setup.
3
Setup
The model is trained on historical generation data and backtested; the result is reported in the dashboard.
Active
A plant-specific 24-hour forecast is published every morning; accuracy is updated as actual generation arrives.

Plant application If you do not have an account, request a demo first.

Sample solar plant· 12.4 MWp · unlicensed

active

Tomorrow

24 hours · MW

Backtest

calculated during setup

Actual

the data you upload

sample card · real plant data is in the dashboard

API integration.

All data in the dashboard is also available through the REST API; with a key it is transferred to ETRM, Excel or in-house models.

All data presented in the dashboard is also available through the REST API. With a key created in the Settings section, forecasts are transferred automatically to ETRM systems, Excel or in-house models.

Hourly series, latest publication, CSV export and accuracy endpoints
Plant forecasts and generation data upload
Read-only keys; account operations only through the dashboard

GET /v1/forecasts/latest

curl -H "X-Api-Key: mgw_…" \
  "https://api.megawattron.com/v1/forecasts/latest?product=demand&country=TR"

200 OK · application/json24 rows

{
  "product": "demand",
  "country": "TR",
  "timezone": "Europe/Istanbul",
  "day": "2026-08-27",
  "rows": [
    { "localTime": "2026-08-27T00:00", "forecast": 31240.5 },    { "localTime": "2026-08-27T01:00", "forecast": 29810.0 },    { "localTime": "2026-08-27T02:00", "forecast": 28930.0 },    …    { "localTime": "2026-08-27T23:00", "forecast": 33180.0 }  ]
}

Organisations we serve.

Energy trading and supply companies

Demand and price forecasts for day-ahead and intraday position management; a direct data feed into portfolio management systems through the API, and project-based portfolio optimisation.

Demand
PTF
API
Portfolio optimisation

Plant owners and operators

Plant-specific hourly generation forecasts for solar, wind and hydro plants; more accurate production schedule (KGÜP) submissions and monitoring of imbalance cost.

Plant forecast
Solar · Wind
CSV

Distribution companies and industrial sites

Demand forecasting at zone, substation or site level; operational planning and battery charge–discharge optimisation in local grids with distributed generation and storage.

Demand forecast
Grid
Battery

Industrial sites with self-consumption

For sites that sell their surplus electricity: their own generation forecast alongside national demand and price forecasts in the same dashboard; internal sharing through the email report.

Plant forecast
PTF
Email report

Demo account and subscription.

Next-day forecasts are published every morning at 10:15. Demo accounts for evaluation are opened by our team; for plan purchases, the subscription is activated the same day once the bank transfer has been confirmed.

Sign-up → plan selection → bank transfer (reference code) → active the same day

Request a demo View plans

Electricity marketforecasts

Our areas of work.

Forecast products

Accuracy measurement.

The daily forecasting process.

Data collection

Modelling

Publication

Access

Verification

Plant-level generation forecast.

Application

Approval

Setup