Load forecasting for procurement: how better demand insight cuts costs

Procurement managers, analysts, and site energy leads can all benefit from better demand insight to inform their load forecasting strategies. Load forecasting introduces a capability that procurement teams often under-invest in, despite its direct effect on pricing and risk. We’ll explore some practical forecasting methods and tools to help gain demand insights and reduce overall costs.

Why procurement performance depends on load accuracy

Load accuracy is important to procurement performance because it clarifies operational outcomes for procurement specialists, including pricing, negotiation ability, and overall efficiency.   

Supplier pricing hinges on forecasted volume and shape

Expected energy production volume is often used as a marker for supplier pricing, as it allows suppliers to estimate the profit they might make at a given level of production and price accordingly. The level of demand, or shape, can also affect pricing because seasonality and expected demand will drive how much a supplier can charge for energy: higher demand can lead to higher pricing, while drops in demand may result in lower energy prices. 

Cost of forecast error 

Because procurement decisions are closely linked to demand, forecasting errors can have a significant financial impact. When the actual volume of energy produced by a supplier does not match the actual energy generated, contractual penalties can be applied. Imbalance pass-through refers to a difference between the submitted expected generation and the actual generation. This occurs when businesses are required to submit their expected energy generation levels in advance of production, typically in specific types of energy markets. These imbalances result in imbalance charges, which the energy producer receives from the operator.

Understanding load shapes and volatility

Load shapes are visual representations of how much energy your business uses over a specific period, with a volatile load shape oscillating up and down and a consistent load shape looking flatter and straighter. 

Baseload vs. peaky profiles

Different power plants are better suited to generate energy for certain types of energy demand. Steadier energy sources, such as nuclear, are usually better suited to baseload profiles, which meet the minimum daily energy demand. Peak load profiles, on the other hand, are more focused on providing energy to meet sudden demand peaks. Natural gas plants are the perfect source to fulfil this need. 

Seasonal and operational drivers

Usually driven by the weather, understanding seasonal changes in summer and winter months can help procurement specialists develop reliable patterns to forecast demand by time of year. Depending on the region, summer can skew demand due to air conditioning needs, while winter can lead to load shape variations due to higher heating costs. Operational drivers are influenced more by the time of day; for example, residential evening spikes or intense industrial activity that occurs at certain times. 

 Identifying predictable vs random demand swings

Sometimes events can cause demand swings. Sudden system failures or market doors are challenging to prepare for, and utilising real-time data allows the market to adapt quickly to sudden swings in demand. When demand swings recur, it is possible to anticipate when they might happen again using historical data which tracks these random cycles: we call these predictable demand swings.

Forecasting methods for corporate buyers

There are several forecasting methods that procurement specialists can use to inform buying decisions. For decisions involving relatively flat, non-volatile products, rolling averages can help calculate average demand over a specific period.

Weather-adjusted models are most suitable for demand affected by weather, for example, seasonal variations that might lead to higher energy use for heating or cooling. We can also integrate machine learning into forecasting; this approach is most suitable for larger energy portfolios.

Using forecasts to improve contract timing and sizing

Systematic over- or under-buying can be an issue in energy procurement because it is difficult to determine energy requirements before a contract is arranged accurately. Forecasting tools that can help avoid over- or under-buying include long-term forecasting, which predicts how a business’s energy needs might change as the company evolves, combined with market research to determine how the energy market itself may evolve. 

Building risk into contracting is a key driver of forecasting, and designing flexible bands in contracts can allow risk tone to be managed more effectively and more structurally. Including tranches in contracts would enable energy procurers to buy parcels of energy in smaller sizes, make contracts more flexible, and allow businesses to react in real time to market changes.

Building a continuous forecasting loop

Forecasting is a valuable tool, but it's crucial to incorporate a process to continuously monitor and refine load forecasts in a continuous forecasting loop. A continuous forecasting loop can examine several relevant factors, such as site feedback and operational change tracking. Energy procurers must also implement a review process to determine how accurate the forecasting was relative to actual energy generated and utilised, integrating a regular review cadence to reflect on the closeness of the forecasting to actual supply and demand. Embedding forecasting into procurecment governance holistically can help drive an organisation's overall strategy. A key tool for all energy procurement specialists, better demand insight doesn’t just improve planning; it directly lowers procurement cost.