Case study


A model will be built to identify behavioural characteristics in electricity usage which may be related to having a vulnerability.


Project VENICE is a customer vulnerability innovation project run by WPD, funded through the regulator innovation funding. The aim of the project is to tackle consumer vulnerability to aid the transition to Net Zero: with the electrification of heat and transport, low carbon technologies are becoming increasingly important, but these technologies are very expensive and people in vulnerable circumstances may be left behind if the correct intervention is not determined now.

Project VENICE aims to tackle this problem in three ways:

  1. Use consumer smart meter data to help identify which consumers may be considered vulnerable.
  2. Determine how the Covid-19 pandemic has impacted the number of vulnerable people, and what behavioural and societal changes will remain.
  3. Determine which interventions are most beneficial to consumers who are vulnerable, to aid them transition to these low carbon technologies. For example, housing insulation or solar panel subsidisation.

Frazer-Nash are leading task 1 to attempt to identify whether a consumer is considered vulnerable from their smart meter data. To do this, we will build a model which identifies behavioural characteristics in electricity usage which may be related to having a vulnerability. Some characteristics we are hoping to identify are:

  • Use of medically dependent equipment in the home,
  • Reduced usage at the end of the month, potentially implying difficulty paying bills,
  • High usage through the winter, potentially implying a poorly insulated house.

The model will be built using a combination of probabilistic, pattern recognition and machine learning techniques. The characteristics we are trying to identify are likely to contain a comparison, for example, lower than average usage for the household type and location. Therefore, one task will be to determine what an average usage looks like for each household type across England and Wales. This will be done by collating LSOA information of housing types and mean usage and combining this with the Acorn geodemographic segmentation of the UK’s population.

To relate the behavioural characteristics identified in the smart meter data to a consumer being vulnerable, we must determine how a vulnerable consumer may use their electricity differently. To do this, our internal team of behavioural physiologists will engage with community groups and charities across England and Wales. This is likely to include, but not limited to: Age UK, British Heart Foundation, Resume Foundation, Citizens Advice, Kidney Care UK, Dementia UK and Mind UK. We will also work with community groups such as Severn Wye and the Wadebridge Renewable Energy Network (WREN) to talk to vulnerable consumers to gain insights directly from them. We will also use this direct engagement to determine how open the public are to having their data being used in this way. 

Apart from the benefit to vulnerable consumers, this project also helps to shape the future regulatory landscape surrounding access to consumer smart meter data for purposes of assistance and planning.


Work with Frazer-Nash

Get in touch and let us help with your next project