Optimal metering plan for measurement and verification on a lighting case study

• A combined spatial and longitudinal optimisation model is proposed to reduce M&V cost.
• The combined optimisation model handles M&V sampling uncertainty cost-effectively.
• The model exhibits a better performance than the separate spatial or longitudinal models.
• The required 90/10 criterion sampling accuracy is satisfied for each M&V report.

M&V (Measurement and Verification) has become an indispensable process in various incentive EEDSM (energy efficiency and demand side management) programmes to accurately and reliably measure and verify the project performance in terms of energy and/or cost savings. Due to the uncertain nature of the unmeasurable savings, there is an inherent trade-off between the M&V accuracy and M&V cost. In order to achieve the required M&V accuracy cost-effectively, we propose a combined spatial and longitudinal MCM (metering cost minimisation) model to assist the design of optimal M&V metering plans, which minimises the metering cost whilst satisfying the required measurement and sampling accuracy of M&V. The objective function of the proposed MCM model is the M&V metering cost that covers the procurement, installation and maintenance of the metering system whereas the M&V accuracy requirements are formulated as the constraints. Optimal solutions to the proposed MCM model offer useful information in designing the optimal M&V metering plan. The advantages of the proposed MCM model are demonstrated by a case study of an EE lighting retrofit project and the model is widely applicable to other M&V lighting projects with different population sizes and sampling accuracy requirements.