Public research showcase for applied Energy Economics and Finnish electricity-market analysis.
FinlandPowerMarketLab is an independent research project examining how electricity demand, weather, generation, prices and power-system conditions interact in Finland.
The longer-term research programme connects:
Demand → Forecast Error → Imbalance → Balancing
This public repository presents the research architecture, selected findings, methods and representative visualisations.
Note: The full research implementation, working datasets, experimental code and development history are maintained separately in a private repository.
FinlandPowerMarketLab currently investigates four connected areas:
- Electricity demand behaviour — how different electricity-user groups behave across hours, weekdays, seasons and temperatures.
- Spot-price formation — how demand, renewable generation, nuclear production and other market fundamentals relate to Finnish electricity prices.
- Demand flexibility — whether consumption shifts away from relatively expensive hours.
- Power-system balancing — how demand behaviour and forecast error may ultimately contribute to imbalance and balancing requirements.
flowchart TD
A["Fingrid & Electricity-Market Data"] --> D["Data Validation & Standardisation"]
B["Electricity Prices"] --> D
C["Weather, Generation & System Variables"] --> D
D --> E["Analytical Research Datasets"]
E --> F["Demand Behaviour"]
E --> G["Spot-Price Formation"]
E --> H["Demand Flexibility"]
F --> I["Demand Forecasting & Forecast Error"]
H --> I
I --> J["System Imbalance"]
J --> K["Balancing Markets & Imbalance Prices"]
F --> L["Econometric Analysis"]
G --> L
H --> L
I --> L
J --> L
K --> L
L --> M["Robustness Checks & Diagnostics"]
M --> N["Research Findings"]
N --> O["Tables / Figures / Public Showcase"]
One research stream studies BE02 (small residential buildings, electrically heated).
An accounting point refers to a separately metered electricity connection.
The analysis found evidence that the relationship between outdoor temperature and BE02 electricity consumption is nonlinear, with a breakpoint around 17°C.
Below the breakpoint, electricity demand falls substantially as temperature rises. Above the breakpoint, the relationship changes.
This supports modelling temperature-sensitive electricity demand with nonlinear specifications rather than assuming one constant temperature coefficient.
The project also investigates which observable market fundamentals are associated with Finnish electricity spot prices.
Variables examined include:
- electricity demand;
- wind generation;
- solar generation;
- nuclear generation;
- cross-border electricity flows; and
- scarcity-related interactions.
Successive model specifications substantially improved explanatory power relative to simpler baselines.
A later multivariate specification reached approximately:
R² ≈ 0.49
with an RMSE of approximately:
€47/MWh
within the analysed sample.
The emphasis is not simply on maximising predictive fit, but on understanding the economic mechanisms associated with Finnish price formation.
The analysis compares the incremental explanatory contribution of different market fundamentals rather than relying only on raw correlations.
This helps distinguish variables that merely move together with electricity prices from variables that add information after other factors are controlled for.
A later research stage separates broader Nordic price movements from Finland-specific price behaviour.
This helps distinguish:
- system-wide market conditions;
- regional fundamentals; and
- Finland-specific price deviations.
The current demand-flexibility research asks:
Do Finnish electricity users shift consumption when electricity becomes relatively expensive within the same day?
The research examines relative within-day prices rather than relying only on absolute electricity prices.
Methods include:
- price-signal correlations;
- relative-price shift tests;
- within-day load allocation;
- high-price versus low-price periods; and
- comparisons across electricity-user groups.
For BE02, the within-day analysis found a negative association between relative price and the fraction of daily electricity consumption allocated to that hour.
Other investigated groups showed different patterns.
These results are treated as observational associations rather than causal price elasticities.
The broader research programme extends beyond electricity consumption.
Electricity demand
↓
Demand forecasting
↓
Forecast error
↓
System imbalance
↓
Balancing requirement
↓
Balancing and imbalance prices
The objective is to investigate whether observable demand behaviour and flexibility can eventually help explain or predict system imbalance and balancing-market outcomes.
Methods used across the research programme include:
- exploratory data analysis;
- multiple regression;
- fixed-effect specifications;
- nonlinear and breakpoint regression;
- event-study designs;
- residual analysis;
- correlation analysis;
- interaction terms;
- standardised coefficients;
- drop-one-variable model comparisons;
- sensitivity testing; and
- robustness analysis.
The research emphasises:
economic interpretation + reproducibility + robustness
rather than model fit alone.
The project uses Finnish electricity-market data, including Fingrid Open Data and Datahub-derived datasets.
The initial research foundation includes datasets covering:
| Area | Example data |
|---|---|
| Electricity demand | Finnish national consumption |
| Electricity production | Finnish generation |
| User-group demand | Distribution-network electricity consumption by user group |
| Customer type | Consumer / company consumption |
| Forecasting | Fingrid electricity-consumption forecasts |
| Renewable generation | Wind-power forecasts |
Additional price, weather, generation and system variables are incorporated according to the research question.