1 INTRODUCTION
Project Objectives
1.1 The central objective of this work was to develop a model to allow the Scottish Government to see the impact of different policy interventions aimed at reducing CO 2 emissions from housing. In particular, the Government wanted to see how it could achieve different reductions targets by 2020 and 2050.
1.2 The Government also had subsidiary objectives for the work, including:
1. Write a critique of the existing BREDEM12 method and data for the purpose of assessing greenhouse gas emissions from housing in Scotland.
2. Draw up a new model specification to assess greenhouse gas emissions from housing, to include the use of lights and appliances and Scottish weather data.
3. Model the changes in housing supply, likely technological advances, and other likely changes in household energy demand.
4. Estimate the costs and benefits of taking up new technologies aimed at reducing CO 2 from housing.
5. Allow the model to be updated and maintained easily so that the Government can model new policies as they develop.
6. Allow users to change all base assumptions in the model.
7. Model the embodied CO 2 and cost of upgrades to new and existing homes.
8. Allow users to focus exclusively on specific subsets of Scottish housing, based on tenure, dwelling type, dwelling age, whether they are 'hard to treat', urban/rural location, or social grouping (pensioners, CERT priority groups, or others).
9. Include in the model the effect of changes in the CO 2 intensity of Scottish electricity over time.
10. Include in the model changes in the cost of different fuels over time, and how this affects savings and cost-benefit calculations relating to the decision to invest in upgrades.
1.3 This was a challenging set of objectives for the model - made all the more challenging by the Government's wish to complete the work in only 10 months. Nevertheless, our model - ' DEMScot' - meets all of these objectives.
1.4 The Scottish Government also asked us to quote to undertake a potential Stage 2 of the model considering the impact of changes in fuel prices and the rebound effect.
Method Overview
1.5 We built a bottom-up model geared to answering specific questions posed by the Scottish Government about how to reduce CO 2 emissions from housing in the years to 2050. The model was created in Excel, and drew heavily on existing research about Scottish housing, CO 2 emissions and economic determinants of carbon emissions - some of it conducted by members of our team.
1.6 There are five components to the model:
1. A building stock/demographics component, based on data 1 provided by Alembic Research and the Scottish House Condition Survey.
2. A building physics component, based on what CAR learnt from developing the LT Method and Climate Lite, and improving on BREDEM-12.
3. A behaviour component, allowing the Government to examine the effect of different households applying factors under user control (such as thermostats, ventilation rates, reliance on electric lighting, and hot water use) differently.
4. Spreadsheets to combine building stock data, performance coefficients and behaviour to enable predictions of the likely impact of policy or information campaigns on aggregate emissions.
5. An embodied energy component that can be revised easily as more reliable data about embodied CO 2 in housing becomes available.
1.7 In addition, Cambridge Econometrics used outputs from this work and their existing MDM-E3 model of the UK economy-energy and environment system to model scenarios representing different policy interventions. These showed the economic impact of different carbon emission scenarios, including costs for the Scottish Government and householders, and the effect on carbon emissions of different macroeconomic scenarios.
Report Overview
1.8 CAR was asked to submit a single report describing all aspects of the project - from the model objectives, to how the model is structured, to the results of scenarios run through the model.
1.9 The report is in eight chapters:
1. Objectives of the work
2. Summary of the literature review which draws out the main findings.
3. Critique of the existing BREDEM12 in order to improve carbon modelling of the housing stock in Scotland.
4. Specification of DEMScot - a technical specification of the model, inputs and outputs, and how the CO 2 and cost calculations function.
5. Economic modelling and considerations - how energy costs and carbon intensities were derived, and broader economics considerations.
6. Critical appraisal of the model - known weaknesses in DEMScot, and pointers for future improvements.
7. Scenarios - the 10 scenarios the Scottish Government asked us to model using DEMScot for their policy work.
8. Moving forward - general thoughts by the team about what this work means for future research and policy relating to housing and CO 2 emissions.
1.10 In addition, there are two appendices. The first is a full description of relevant parts of our literature review. The second is a summary description of the Expert Panel meeting we ran last year.