Energy Economics and Econometric Demand Forecasting Training Course

Course overview

Energy businesses must make investment, procurement, hedging, tariff and capacity decisions before demand is known. A forecast that ignores weather, prices, economic activity, customer mix or structural change can create expensive consequences: over-contracted fuel, avoidable imbalance charges, stranded capacity, missed revenue and weak regulatory submissions. Professionals need to distinguish a plausible demand narrative from a statistically defensible forecast, explain uncertainty to decision-makers and identify the assumptions that materially change the result.

This course develops applied energy economics and econometric forecasting capability for electricity, gas and related energy markets. Participants examine demand drivers, price elasticities, load shapes, seasonality, weather normalisation, industrial and residential customer segmentation, and policy effects. They build and test regression, time-series and econometric demand models using Excel, EViews and Python; assess stationarity, autocorrelation, multicollinearity and forecast error; and compare baseline, scenario and sensitivity forecasts. The programme also addresses the economic interpretation of model outputs for tariffs, investment cases, procurement plans and regulatory reporting.

Teaching combines concise instructor-led explanations with worked energy-market datasets, model-building laboratories and decision-focused case discussions. Participants estimate a demand equation, validate it against holdout data, document assumptions and convert results into an executive-ready forecast pack. Each participant leaves with a reusable demand forecasting workbook or notebook, a model specification template, diagnostic checklist, scenario framework and a practical plan for applying the method to their own organisation's data.

The course is designed for analysts and managers who already work with energy, financial or operational data and need stronger forecasting discipline rather than a purely theoretical econometrics programme.

Course objectives

By the end of this course, participants will be able to:

• Specify an energy-demand model linking consumption to price, income, weather, calendar effects and customer segments
• Estimate price and income elasticities using log-linear regression models
• Build seasonal time-series forecasts using ARIMA and seasonal dummy-variable methods
• Test demand data for stationarity, autocorrelation, multicollinearity and heteroscedasticity
• Validate competing models using holdout samples, RMSE, MAE, MAPE and residual diagnostics
• Construct weather-normalised demand forecasts from heating and cooling degree-day data
• Create baseline, high-demand and low-demand scenarios with documented economic assumptions
• Produce an executive forecast pack containing model outputs, uncertainty ranges, risks and recommended actions

Benefits of attending

Target competencies

Who should attend

• Energy Market Analysts — who forecast electricity or gas demand for trading, balancing and commercial decisions
• Energy Economists — who need to quantify demand drivers and communicate elasticities credibly
• Demand Forecasting Analysts — who want to strengthen model diagnostics and scenario design
• Utility Planning Managers — who require evidence for capacity, network and resource planning
• Energy Procurement Managers — who need demand outlooks to support fuel, power and hedging commitments
• Financial Planning and Analysis Professionals — who translate energy-volume assumptions into budgets and investment cases

Requirements and prerequisites

Participants should be comfortable working with tabular data in Microsoft Excel, including formulas, charts, sorting and basic pivot tables. The course assumes familiarity with descriptive statistics, percentages, rates of change and the business meaning of energy demand, tariffs and consumption data. Prior exposure to regression is useful but not essential; key concepts are refreshed before model estimation. Participants should bring a laptop capable of running Excel and either EViews or Python. No advanced calculus, prior coding expertise, data-science qualification or previous use of econometric software is required.

Training methodology

The five days alternate between instructor-led economic interpretation, guided software demonstrations and hands-on modelling labs. Participants work with realistic electricity and gas demand datasets containing weather, price, calendar and macroeconomic variables. Short cases examine forecasting decisions in utility planning, energy procurement and tariff setting, while group reviews focus on challenging assumptions and interpreting diagnostics. Each day closes with a practical output, progressing from data preparation to a tested model and scenario pack. On the final day, participants adapt the workflow into an application plan for their own forecast cycle.

Course outline

Day 1: Energy demand economics and data design

• Energy demand, load and consumption measures
• Short-run and long-run price elasticity
• Income, output and demographic demand drivers
• Residential, commercial and industrial customer segmentation
• Weather variables and degree-day construction
• Energy demand data sources and data-quality checks
• Forecast purpose, horizon and granularity definition

Workshop: Participants map the drivers, data fields, forecast horizon and decision use for a utility demand-forecasting case and produce a model scoping sheet.

Day 2: Regression models for energy demand

• Linear and log-linear demand model specification
• Ordinary least squares estimation in EViews and Excel
• Interpreting coefficients, elasticities and marginal effects
• Seasonal, weekday and holiday dummy variables
• Lagged price and income variables
• Confidence intervals and hypothesis testing
• Model specification choices for electricity and gas demand

Workshop: Participants estimate and interpret a log-linear electricity-demand regression, then produce an elasticity summary for commercial stakeholders.

Day 3: Time-series forecasting and diagnostics

• Trend, seasonality and cyclical demand patterns
• Stationarity testing with augmented Dickey-Fuller tests
• Autocorrelation and partial autocorrelation analysis
• ARIMA and seasonal ARIMA model selection
• Residual plots and Ljung-Box autocorrelation tests
• Multicollinearity and variance inflation factors
• Heteroscedasticity testing and robust standard errors

Workshop: Participants compare regression and seasonal ARIMA forecasts on a holdout sample and produce a diagnostic decision note.

Day 4: Forecast validation, weather and scenarios

• Training, validation and holdout sample design
• RMSE, MAE, MAPE and bias measurement
• Weather normalisation using heating and cooling degree days
• Outlier treatment for outages, lockdowns and abnormal events
• Baseline forecast construction and assumption registers
• High-demand and low-demand scenario design
• Sensitivity analysis for price, weather and economic growth

Workshop: Participants build three weather and macroeconomic demand scenarios and produce a forecast range with an assumptions register.

Day 5: Decision-ready forecast packs

• Translating volume forecasts into procurement implications
• Demand forecasts for tariff and revenue planning
• Forecast uncertainty and risk communication
• Forecast governance, version control and audit trails
• Dashboard charts for forecast versus actual performance
• Python pandas and statsmodels forecasting workflow
• Executive presentation of model limitations and recommendations

Workshop: Participants complete an executive-ready demand forecasting pack containing a tested model, scenarios, accuracy metrics, risks and recommended actions.

Tools & standards covered

Microsoft Excel, EViews, Python, Jupyter Notebook

A typical training day

Live online deliveries follow the same structure in the East Africa Time zone, with shorter screen blocks and longer breaks.

What the fee includes

How you can take this course

Certification

Participants who complete the full five days receive the Skillset Development Certificate of Completion, stating the course title, course code, dates and delivery format — suitable for professional-development records and employer reimbursement.

Frequently asked questions