SAVE Customer Model

Presentation on the SAVE Customer Model developed by the University of Southampton and delivered during the closedown event at Central Hall, Westminster on 6th June 2019.

Transcript

1. June 2019 SAVE Close Down Event Customer Model
2. Customer Model Content 2 • Project objectives • Customer Model inputs: SAVE data streams • Customer Typology • Customer Model outputs: • ‘Baseline’ demand profiles • Intervention impact profiles • Summary
3. Customer Model Part of the Network Investment Tool 3 Customer Profiles Customer Model Network Model Pricing Model Network Data Census Data Investment Decision SAVE Intervention Feasibility Investment Strategy Future Scenarios Intervention Costing The Customer Model provides customer load profiles to the Network Model
4. Customer Model Project objectives 4 • Analysis of the household characteristics that capture diversity in load and treatment effects in trial groups; • Ability to produce ‘baseline’ half-hourly electricity consumption profiles at the individual household level; • Produce similar profiles for trial intervention groups taking account of intervention trial effects; • Produce similar profiles for designated Census areas in the Solent region; • Estimate the change in electricity consumption at specific times of day that can be attributed to the SAVE intervention trials;
5. Customer Model Model input data 5 Q1 Q2 Q4 Q3 Q1 Q2 Q4 Q3 2017 2018 Control TG1 Treatment TG2 Treatment TG4 Treatment TG3 TP1 TP2 TP3 Treatment Group Trial Periods Monitored Electricity demand (Wh/15-mins) 4,000 households over 2 years Household surveys: • Data collected at recruitment • Socio-demographic characteristics • Dwelling characteristics • Appliance ownership
6. Customer Model Customer Typology 6 Q1 Q2 Q4 Q3 Q1 Q2 Q4 Q3 2017 2018 Control TG1 Treatment TG2 Treatment TG4 Treatment TG3 TP1 TP2 TP3 Treatment Group Trial Periods Variable selection Customer Type definition Census small area statistics Customer types: • Developed using data from all groups prior to interventions; • Defined by household characteristics available in Census small area statistics; • 3 characteristics selected that best predict load during peak hours; • Characteristics vary across small areas providing a different mix of customer types for each small area.
7. Customer Model Customer Typology 7 HOUSEHOLD SIZE (PERSONS) 4 CATEGORIES: - 1 person - 2 person - 3 person - 4+ person Image from http://datashine.org.uk (UCL)
8. Customer Model Customer Typology 8 DWELLING SIZE (BEDROOMS) 4 CATEGORIES: - 0-1 bedrooms - 2 bedrooms - 3 bedrooms - 4+ bedrooms Image from http://datashine.org.uk (UCL)
9. Customer Model Customer Typology 9 PRIMARY HEAT SOURCE 3 CATEGORIES: - GAS - ELECTRIC - OTHER Image from http://datashine.org.uk (UCL)
10. Customer Model Model outputs – ‘baseline profiles’ 10 Q1 Q2 Q4 Q3 Q1 Q2 Q4 Q3 2017 2018 Control TG1 Treatment TG2 Treatment TG4 Treatment TG3 TP1 TP2 TP3 Treatment Group Trial Periods Baseline profiles TG1 Variable selection Customer Type definition CUSTOMER TYPES
11. Customer Model Model outputs – ‘baseline’ demand profiles
12. Customer Model Model outputs – ‘baseline’ demand profiles Electrically-heated households Other non-gas heated households
13. Baseline profiles TG1 Customer Model Model outputs – intervention profiles 13 Q1 Q2 Q4 Q3 Q1 Q2 Q4 Q3 2017 2018 Control TG1 Treatment TG2 Treatment TG4 Treatment TG3 TP1 TP2 TP3 Treatment Group Trial Periods Variable selection Customer Type definition Intervention impact TG3 Intervention impact TG4 Intervention impact TG2 CUSTOMER TYPES
14. Customer Model Model outputs – intervention profiles
15. Customer Model (CM) Profile generation for scenario e.g. LEDs (TG2) Network Model (NM) ‘Customer Type’ demand and intervention profiles Census Interface x6 x1 x4 x17 x45 x12 Feeder 85 households (all gas-heated) Customer Model Model interaction
16. The Customer Model provides: • A Customer Typology … • Representing greater diversity of demand, and • aligned to Census data to mapping of profiles to Census output areas; • ‘Baseline’ half-hourly electricity consumption profiles • Intervention impact profiles providing estimated change in electricity consumption Customer Model Summary
17. Thank you for listening. t.w.rushby@soton.ac.uk @tom_rushby #SAVEClosedown Customer Model