This is a new model mainly for segmenting metered time series into end-use or different customers. It is based upon load curves from national load research projects. The model uses statistical methods and handles climatic dependencies and the diversification in the load from different customers. It can also estimate the coincident peak demand in a network with selected degrees of confidence.
USELOAD’s origin is from the international liaison body EDEVE where load-research, demand side management (DSM) and more specific consumption relations are being discussed and elucidated. A general need for more detailed description and modelling of end-uses resulted in a development-project with SINTEF Energy Research as the executive part.

The speciality of USELOAD is great flexibility, basic development of methods and great applicability for different kinds of purposes. Detailed input data is important before the model is operative for a specific region. Typical daily load curves for i.e. lighting, heating, ventilation, hot water and other electrical appliances should be established. Based on these data, the total load curve can be calculated divided into hours for a day or i.e. for a year. Table 1 shows a random example of such results from a specific supply region/transformer with a specific division in households and industry. All levels from one single house and up to e.g. total Norway can be modelled in a very user-friendly way – dependent of the number of single loads or buildings in different customer categories (houses, offices, schools, social etc.)

Table 1: A random example: Coincident peak demand is 11711,4 kW at 09:00 a.m. in week 7. The local peak for household customers is metered at 8,6 kW (95% confidence interval). The lowest temperature is –13 °C.

Examples of results from Useload. Graphic results produced with the use of Useload shows many interesting relationships. Two examples of graphs are shown below: Segmentation of total load into different customer groups, and segmentation of total load into different end-uses (purpose of use).

The results of the calculation and the possibility for consequence analysis in USELOAD are a good basic for different kinds of analysis. Useload can be an interesting tool in a number of applications:

• The resulting demand of changing to new energy efficient technology i.e. high efficiency lighting
• Calculation of power segmented into different end-uses in different supply regions
• The effect of different kinds of energy control
• Detailed energy and power forecasts at all voltage levels
• The effect of changing to energy flexible solutions
• Detailed classification and mapping of load in different parts of the network where the maximum power or other customer relevant data are available
• Climate correction of energy and power with hour’s division
• Detailed calculation of network losses

Utility value for different needs
The model constitutes as an interesting tool for different parties in the power market. This can shortly be described as following:

Network owner: Net planning, energy planning, design of tariffs and analysis of customers
Power supplier: Market analysis, development of energy services and new products
System operator: Power analysis and forecasts
Authorities: DSM analyses and design of policy instruments towards customers

An interesting observation is that all the six involved countries focus on the challenges a deregulated power market will bring to better understand and model the energy market.