National Physical Laboratory

New NPL model optimises smart technology for the electricity grid


Over the past 50 years, electricity consumption in the UK has increased by some 180 % and is set to grow further with increasingly electrified heating and transportation. At the same time, a fifth of the UK's power stations are due to close by 2020 due to aging infrastructure and parallel technological development, requiring £110bn of new investment to replace the stations and upgrade the National Grid.

As electricity generation accounts for around a third of the UK's CO2 emissions, the government is keen to lower its impact and has a target to cover 20 % of the UK's energy demand from renewable sources by 2020 next to legally binding emissions reduction targets (80 % by 2050). While renewable sources are generally CO2 neutral, their intermittent supply, coupled with the current lack of energy storage capabilities, results in a significant challenge in balancing irregular supply with consumer demand.

Efficient investment in smart interventions

In the future, smart infrastructure and smart technology will help manage the electricity grid. An ICT-enabled power grid will allow information to be exchanged across the system to deliver an integrated management of supply and demand, allowing intermittent renewable supply to be redirected efficiently to areas of demand and in doing so minimise carbon emissions.

Significant investment is required to upgrade the existing infrastructure to enable smart intervention on power networks. Assessing which investments will provide the best return on investment in terms of emissions reductions requires reliable calculations of the difference in carbon emissions between business-as-usual and the post smart intervention, optimised scenarios. This will allow regulators and investors to implement the most effective and efficient methods.

NPL has developed an independent, adaptable and tested methodology to model the amount of carbon savings in electricity generation and consumption, as well as other aspects of the smart infrastructure, to provide confidence to investors and maximise the economic, societal and environmental impact of investment in the smart grid.

The NPL model represents the electricity grid as a mathematical simulation consisting of continually changing inputs and outputs, accounting for generation methods and consumption patterns that vary depending on conditions and demand. It is based on Ensemble Kalman Filter (EnKF) forecast and optimisation (EnOpt) routines, and uses real-world electricity generation and consumption data, to tune, test and refine the output. The resulting estimation of carbon savings is provided through a statistically robust and rigorous methodology with uncertainty quantification that has been tested on a number of case studies.

Case study: Nesta Dynamic Demand Challenge Prize

The Dynamic Demand Challenge Prize, run by Nesta's Centre for Challenge Prizes in partnership with NPL's Centre for Carbon Measurement, set out to find new ways of shifting electricity use to off-peak times or towards times of high renewables generation - a method known as 'Demand Side Response' (DSR). Teams with innovative DSR solutions entered the competition to have their products tested for potential impact, with the winner receiving funding to further develop their concepts.

Each of the five finalists had their innovations assessed using NPL's Carbon Savings Model to determine how effectively these reduced carbon emissions. The model used a common scenario to allow the comparison of these different DSR approaches. Using NPL's analyses, the winner - Hestia powered by Demand Shaper - was found to have a potential annual carbon savings of 2.8-3.8 tonnes of CO2 per installation. If scaled-up, by 2020 the innovation could provide a peak demand shifting capacity of 1.7 GW, potentially saving up to 4 million tonnes of CO2 per annum. NPL's model enabled the finalists to be evaluated through a quantifiable and comparable method and provided a measure of impact of the prototypes that the finalists produced.

Case study: Demand Side Response (DSR) of a UK Energy Aggregator

As outdated power stations close, the margins between the level of generated and needed power tighten, adding particular pressure during periods of peak demand on cold winter evenings. To meet demand, either power supply has to increase through expensively ramping up reserve capacity at power stations that tend to be emissions intensive, or demand has to be curbed. This can be done by having electricity users, including businesses, temporarily switch to local backup generators.

This type of DSR programme can employ diesel generators to cover these periods of high demand. While diesel is among the most polluting fuels, it can be more efficient than coal or gas power stations for short and limited periods of generation to meet the peak demand.

NPL has worked alongside a major UK company to calculate the reduction in carbon emissions - the 'carbon savings' - through their participation in such DSR programmes. Using the NPL model, it was found that all 51 self-owned generator sites used by the company showed a reduced amount of carbon emissions compared to purchasing power from reserve capacity power plants.

The model also calculated the critical duration of using the diesel generation, above which it was more carbon-efficient than use of the reserve power stations. This provides an estimation of the most optimal number of generators and duration of their operation to balance the required power supply in the most environmentally friendly way.


The UK environmental and low carbon market was valued at £128bn in 2012 and is expected to have grown by 25 % by to this year. It is critical that we ensure investments in low carbon innovation are made with confidence in the knowledge that the resultant technologies will deliver the maximum reduction in carbon emissions required for the UK to meet its legally binding emissions reduction targets. NPL's Carbon Savings Model provides a robust and tested tool to assess the savings being achieved and can ensure consistent measurements to allow comparisons between different methods.



For further details, please contact Marieke Beckmann or Valerie Livina

Last Updated: 30 Sep 2015
Created: 14 Sep 2015


Please note that the information will not be divulged to third parties, or used without your permission