Oil & gas load forecasting & demand response

This data app demonstrates a compact data app that forecasts short‑ to mid‑term load, and identifies when/where demand response (DR) can shave peaks most cost‑effectively for oil & gas facilities. It also plans DR events and incentive levels to meet capacity constraints while minimizing curtailment costs to customers.

System load vs capacity

Forecasts short‑ to mid‑term load and identifies when/where DR can shave peaks most cost‑effectively.

Site‑level load & DR potential

Plans DR events and incentive levels to meet capacity constraints while minimizing curtailment costs to customers.

Short‑term forecasting

We train a simple RandomForest on time, weather, price, and lag features. We use the first 10 days for training and the last 4 days for testing.

Identify DR opportunities

We compute hours where the predicted system total exceeds the capacity limit and rank site‑hour opportunities by program cost ($/MWh) subject to each site's minimum incentive to accept curtailment.

Scheduled DR (by hour & site)

KPIs & business impact

We compute key program metrics and avoided emissions (assuming grid emissions intensity in the dataset).