Our work

Access to cooling is essential for meeting our social and economic goals but equally unmanaged growth in cooling represents one of the largest end user threats to achieving our climate goals for CO2 emissions. To address this, we urgently need access to clean cooling for all.
Air conditioning units on the outside of housing

Read our A Cool World: Defining the Energy Conundrum of Cooling for All report (PDF)

Our work to date work, including research with stakeholders across the sector, has led to a clear set of recommendations.Given the urgency and magnitude of the challenge and the multi-partner and multi-disciplinary research and delivery mechanisms required, to lead this work we urge the establishment of a multi-disciplinary Centre of Excellence for Clean Cooling (CEfCC) to bring the global expertise together to research and develop the step-change pathways (culture and social, technology, policy, business models, financing) for achieving (i) cheapest cost (whole of life), (ii) greatest energy system resilience and (iii) lowest carbon emissions while (iv) meeting social and economic cooling needs. To this end we are already working with a series of partners from academia, research institutes, Governments, industry and NGOs.

What is "Clean Cooling"?

Meeting our cooling needs sustainably within our climate change, natural resource and clean air targets. Clean cooling necessarily must be affordable and accessible to all to deliver the societal, economic and health goals. It likely starts with mitigating demand.

What needs happen to deliver Cooling for All sustainably?

Delivering sustainable Cooling for All
 Roadmap Delivery Accelerate
All-stakeholder engagement
Engage and drive collaboration across the main stakeholder groups (policy, customers, industry, developers and financiers)
Fund Innovation development
Connect research institutes OEMs, VCs, policy makers and customers to collaborate on the delivery of high impact innovation.
 Policies to unlock finance
Create the market environment (policies and business models) to attract infrastructure investment to deliver "Cooling for All"
Systems Level Analysis
Assess Cooling for All at the systems level - size of the challenge and alternative technologies, energy sources, business models and cross-industry resource efficiency sharing mechanisms.
Eliminate the performance risk and demonstrate impact through live market testing and validation in Living Labs
Identify the skills gap (design through to installation and maintenance) and  connect educational institutes OEMs,  policy makers and customers to collaborate on the delivery of accelerated solutions
Create the Intervention roadmap (technology, policy, finance, etc) to deliver 70% reduction in electricity usage for cooling.
Design manufacturing processes and engage industry to scale novel technologies; ideally using a global science, local delivery model
Effective Knowledge Transfer
Use system level model, in-country living labs and manufacturing accelerator to roll out "fit for market" solutions across new geographies

Unintended Consequences

Identify, plan for and mitigate potential unintended consequences.


Research programmes

Energy storage

Professor Yulong Ding is a world expert in thermal energy systems and leads the Birmingham Centre for Energy Storage. He worked with Professor Toby Peters to develop the technology behind the Highview Liquid Air Energy Storage and is leading international projects such as the development of novel air conditioning systems for high speed rail.

Dr Jonathan Radcliffe is an expert on the modelling of the integration of energy systems and with Professor Peters is involved the in £7M EU Cryohub project. Dr Radcliffe has also developed Birmingham's Masters programme in Energy Systems.

Dr Raya Al-Dadah and Dr Saad Mahmoud are experts in thermal engineering and are working on the development of novel adsorption cooling systems.

Cold and Power power

Professor Thanos Tsolakis and Dr Karl Dearn are leading on Cold and Power and are working with Dearman on the development of its cryogenic engine for advanced, zero-emission cooling in developing markets.

Sustainable energy use

Professor Peter Fryer is an expert in food technology and is a co-director of the UK Centre for Sustainable Energy Use in Food Chains.

Dr Rosie Day is an expert in energy social science and Co-Investigator of The DEMAND Centre and together with Professor Peters plans to look at the unintended consequences and the impacts of introducing a major shift to dynamic socio-technical systems. The University of Birmingham also leads the UK’s Energy Research Accelerator which includes developing the novel Industry 4.0 manufacturing processes (including Factory in a Box) to move new cooling technologies through to market.

Unintended consequences of cooling

Introducing more affordable and readily available means of cooling in food supply chains and the built environment is not just a matter of adding cooling to the status quo; it will introduce major shifts to dynamic socio-technical systems as well as the wider environment and eco-systems. These could result in a number of unintended and sometimes negative, as well as positive, effects. It is important to try to identify and plan for these in advance.

For example:

  • A cold chain could allow farmers to transition from staple to high value (but temperature sensitive) horticulture. A move to potentially more water demanding produce could have implication for water resources.
  • The provision of food supply chain cooling will allow farmers to reach more distant markets. More processing at the farm could lead to increased packaging demands, in itself a a major source of environmental pollution
  • Refrigeration in the home can change cooking styles and patterns – especially the case if coupled with more processed food and the convenience products that cold chains enable. This can affect can indigenous diets and health. Domestic refrigeration can also reduce the frequency of shopping which can affect local marketplaces.

These are but a small number of examples, yet they illustrate clearly the important of research work to identify potential unintended negative social, ecological or economic consequences and engage to mitigate them as soon as possible.

Domestic Air Conditioning in 2050

Jenny Crawley, Stephanie Ogunrin, Shivani Taneja, Inna Vorushlyo, Xinfang Wang


The latest projections from the Met Office show that the UK annual average temperature is set to increase by 0.7-3 degrees from the 1981-2000 to the 2041-60 period (Met Office, 2019). Alongside average increases, summer temperatures will rise, as well as the frequency of heatwaves. The Met Office estimate that the extreme hot summer experienced in 2018 would be less than 0.5% likely with no manmade climate change, is 12% probable currently, and will be 50% likely by 2050 (Met Office, 2018). This raises the question of the likely uptake of domestic air conditioning.

Previous cooling scenarios and assumptions

Little data exists for the UK quantifying the relationship between AC uptake and its predictors. National Grid (2019) assumed a 60% penetration by 2050 (based on all homes in urban areas installing air conditioning). The Tyndall Centre (2016) constructed projections for 2030 using various scenarios. One was based on the current growth of the market (1.6% by 2030), one was based on growth as seen in other countries (2.9% by 2030), and two were based on simple DECC scenarios (33% and 67% by 2051). Finally, Peacock et al (2010)  used the relationship between uptake and cooling degree days observed in America to predict 18% uptake in London by 2030, caveating this finding due to behavioural differences between the U.S. and England.

Aims of this project

  • To construct a set of socio-technical scenarios for 2050 AC penetration
  • To provide a first estimate of grid impacts based on temporal characteristics of air conditioning use
  • To identify the data gaps and areas for future collaboration

Scenario construction

The aim was to produce a small number of scenarios using a limited set of social and technical variables which are known from the literature to be important predictors of air conditioning uptake. In an ideal world, an uptake model from now to 2050 would have been constructed but the data for this were not available, hence using a scenario approach. We focus on England to narrow the scope.

Building simulation was carried out using UCL’s HPRU model to explore some physical variables: geographical location, building age and dwelling type. Literature and other theory was used to examine the effect of other variables: heat stress, children, income, tenure, urban heat island.

Four scenarios

Four narratives were constructed as follows:

  1. Building-based scenario: homes built between 1990 and 2025 overheat more than others and all occupants install air conditioning. After 2025, the Future Homes standard for new buildings kicks in and overheating is significantly mitigated.
  2. Age demographic (1): all households where the Household Reference Person is aged 75 or over install get air conditioning. This takes place across the socio-economic and tenure spectrum, perhaps as a result of a policy equivalent to the Winter Fuel Discount which only considers age.
  3. Age demographic (2): all owner-occupied dwellings with at least one dependent child install air conditioning.

  4. Wealth demographic scenario: all owner-occupied dwellings with income above the median install air conditioning. 

Since climate is known to be an important factor (Yun & Steemers, 2011), each of the above four scenarios had two implementations: one for all dwellings in the category, and one confined to urban environments in the south of England. ‘Urban’ was defined as the three ‘predominantly urban’ categories of conurbation defined by DEFRA (2011).

Summary of scenario results

  • The range of air conditioning penetration from the above scenarios was 5-32% of English households
  • This is obviously a very simplistic treatment of uptake which uses crude and simple categories and some combinations of categories. In reality, within a given category, not ‘all’ the population of that category would adopt air conditioning.
  • All of the scenarios resulted in less than a third of households adopting AC. This is much lower than National Grid’s prediction of 60%.
  • They have been constructed so that the location/weather variable (the south/urban flag) makes a big difference (5-12% of households if the flag is applied, 19-32% if it is not). The size of this difference in reality is currently unknown in England.

Building Modelling

In order to estimate the grid impacts of the above scenarios, the half hourly AC power consumption on a hot day was estimated for two of the most common building types in the UK. Therefore, a one-storey semi-detached house and a three-storey block of flats was modelled using Autodesk Revit. Both dwelling types were modelled using standard current standard construction.


It was intended to carry out building simulation using 2050 climate projections, however the free Prometheus files available displayed erroneously cooler weather than currently so were regarded as not compatible with latest Met Office predictions (Met Office, 2019b).  Therefore, current weather files for London Heathrow were generated using Meteonorm data. Designbuilder software was then used to carry out half hourly dynamic thermal simulations of the 2 buildings using the weather files.

Key features of the building modelling are summarised here:

  • Timings of air conditioning use were based on the best available data: the UK-based empirical study by Pathan et al (2008) and some additional building simulation showing the time of the warmest internal temperatures to be 2-6pm
  • Windows were assumed to be closed at night, in accordance with Mavrogianni et al  (2017) who showed that this was the case in a proportion of London homes for reasons of noise and security
  • Internal temperature above which AC was switched on was also based on Pathan et al (2008) as 25°C for living rooms during the day/evening and 20°C for bedrooms at night.

Electricity System Modelling


  • The AC loads derived in the previous section were simply multiplied by the uptake scenarios above to give a total AC load. A COP of 3 was used to convert kW(th) to kW(elec)
  • National Grid’s ‘Two Degrees’ scenario (National Grid, 2019) was used to predict the other demands on the grid, as well as the solar and typical wind power. The other grid demands included significant EV charging.
  • AC load was then added to the other grid demands to investigate the potential effect of air conditioning. 


Domestic Ait Conditioning in 2050 Project Results Graph

Our highest (i.e. worst case) scenario increases the evening peak by 7 GW. In this scenario the reason for the peak increase is the pre-existence of an evening EV charging peak coincidental with AC demand. AC is not coincidental with solar PV generation.

This is based on scenario modelling and not absolute truth, therefore the best interpretation of the results is perhaps to explore the issues and questions it raises:

  • If EV charging and AC are coincidental, does this present a problem for the electricity system? If so, which demand should take priority and which should be shifted?
  • How can renewable generation be reconciled with AC demand which occurs hours later?
  • What is the effect of AC demand diversity on the results? 

Project Summary 

  • Domestic cooling is an energy demand meriting more investigation
  • If air conditioning is used as we assume in our scenarios, summer peak will increase and demand will not coincide with optimum renewable generation
  • Key gaps in the evidence base were found to include:

 What motivates English households to install air conditioning? What time of day or night is air conditioning really used? How many rooms will air conditioning be installed in per home? How coincident is the AC demand across UK households?