This document summarizes a PhD thesis on developing a spatially enabled model to estimate energy consumption in sub-city areas in Newcastle upon Tyne, England. The model estimates electricity and gas usage at multiple scales from districts down to individual properties. It analyzes factors like housing types, insulation levels, and socioeconomic characteristics that influence energy use. The model is evaluated against national energy estimates and shows varying accuracy depending on the homogeneity of the area studied. Insights from the analysis can help inform local energy and climate policies.
Digital Commerce Lecture for Advanced Digital & Social Media Strategy at UCLA...
Esriuk_track8_newcastle_university_spatial
1. Javier Urquizo
PhD Candidate
19th May 2015
Acknowledgement:
Adrian McLoughlin, Climate Change Officer Newcastle City Council.
Supervision:
Carlos Calderón, School of Architecture, Planning and Landscape, Newcastle University
Philip James, School of Civil Engineering and Geosciences, Newcastle University
Peter Kellett, School of Architecture, Planning and Landscape, Newcastle University
2. • Problem description.
• Aim and objectives.
• The energy model description.
• Individualized data.
• Aggregated data (MLSOA / LLSOA).
• The spatially-enabled database.
• Descriptive statistics.
• The Case study areas: MLSOAs, LLSOAs.
• Energy (Electricity and gas) area based estimation at different scales and discussion.
– Simplified model results – districts.
– Detailed model results – neighbourhoods.
• Heating gas consumption estimation in individual properties and discussion.
– Retrofit measures
• Spatial distribution of energy intensity and patterns of dwellings with similar energy profiles.
• Fuel poverty.
• Sensitivity analysis.
3. • Cities need to contribute to energy and climate policy goals:
− 80% GHG emission reduction by 2050.
− CCC fourth carbon budget: solid wall insulation and heat
pumps (Committee on Climate Change, 2013).
− For fair energy efficiency policies. Should be tied to
resources which enable the fuel poor to reduce energy
consumption.
• Building retrofit, district heating, LCZ.
• Local Strategic Partnerships (LSP) important role to area-
based initiatives (Social Exclusion Unit, 2011, p. 56).
• Need robust evidential base for decisions:
– Quantitative information (understanding energy use)
– Informing strategic interventions.
4. Aim
The aim of this research is to develop a spatially enabled model to estimate the
energy consumption in sub-city areas in a multi-scale approach.
Objectives
Develop framework for domestic energy (gas and electricity) in Newcastle upon
Tyne. Explaining the local area characteristics as drivers of energy in three
scales:
◦ Simplified district model - (MLSOA).
◦ Detailed neighbourhood model – (LLSOA).
◦ Retrofitting model - Geographic Area: from a single dwelling to community.
Spatial operators: Precise spatial extent of the energy consumption sub-city
areas
Reverse lookup procedure: Identifying building aggregated areas similar spatial
expression patterns. Fuel poverty.
Assess the efficacy of the energy model with respect to sensitivity.
5. Modelling methodology approaches at different scales.
• The simplified (cluster) model.
− Modelling approach is top-down;
− Use of the two-step sequential cluster: first identify the cluster structure
(hierarchical cluster) and second optimal cluster method to assign observations (k-
means cluster analysis);
− Energy dwelling archetype prototypes in clusters. Capture the mean effects in the
interaction between the physical variables
The result is a simplified cluster model with a medoid building prototype.
The detailed model.
– Modelling approach is bottom-up;
– Individual buildings;
– Cohesive energy structures. RSL, HMO, E7, district and group heating.
Unknown information at known locations in a geographic area to complete input to
energy model.
− Point interpolation (data is collected at point locations): Nearest neighbour, Inverse
distance weight, Kriging;
− Converting data between geographies. Multiple imputations
The result is a detailed spatial micro-scale energy model (at property level)
6. • OS Mastermap building outlines.
− 182,110 features – Toids.
• Local authority Property Gazetteer.
– 122,733 Residential dwellings – UPRN.
• Cities Revealed building age and type - Building class.
• Warmzone: households that presumably are in fuel poverty (EST
et al., 2005).
– 78,163 household surveys – 63.69% coverage
• English House Condition Survey
– Survey of around 16,150 households (including vacant
dwellings) with a follow up physical inspection on 15,523 cases
where an interview with the household was also secured, full
SAP and energy profiles;
– Used by DECC to estimate national energy demands.
7. ID Sub-theme Unit Indicator Source
1 Demographic Population Population density Census
2 Demographic Households Household composition Census
3 Demographics Households
Number of People Living in
household
Census
4 Demographics Households Tenure Census
5
Economic
structure
Households Income band Census
6
Economic
structure
Socioeconomic Classification
Experian
data set
9
Local
Environment
Building Number of rooms Census
10 Social Households Fuel poverty Census
8.
9. • Interesting observations on the descriptive of the housing
stock:
– Cavity wall insulation: 82% have cavity walls. The un-
insulated in excess of 50%;
– Most of solid walls are un-insulated 50.8 %. However,
high capital costs;
– Double glazing: 66.3%;
– Electric provision of heat is 41% in Westgate LLSOA
8440, and 76% in Westgate LLSOA 8397. One possible
solution is provision of LZC technologies;
– Opportunities for off-grid generation, using a diverse and
independent renewable energy provided on-site in Castle.
10. • Interesting observations on the descriptive measures of
the housing stock:
– Tenure: LA 28.9%; Private rented 11.0%; Owner
Occupied 56.1%;
– Heating System: Combinational boilers 47.3%; Standard
Boiler 36.2%; Condensing Boiler 1.8%; Condensing
Combinational 2.1%;
– Heat fuel: gas 91.4%; E7 6.9%;
– Cook fuel: gas 74.4%; electric 24.9%;
– Age: 91.8% pre-1980.
11. @ Crown copyright. This
research uses 2001 MLSOA
and LLSOA data provided
trough EDINA UK Borders with
the support of ERCS and JIST
and uses boundary material
which is copyright for the crown
12.
13. 𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 = � 𝐸𝐸𝐸𝐸 𝑥𝑥 𝑁𝑁𝑁𝑁
8
𝑖𝑖=1
MLSOA
Annual
Electricity
consumption
(KWh)
Annual Gas
consumption
(KWh)
Total (KWh)
Annual
Electricity
consumption
(KWh)
Annual Gas
consumption
(KWh)
Total (KWh)
Annual
Electricity
consumption
(%)
Annual Gas
consumption
(%)
Total (%)
Castle 16,213,837 55,654,385 71,868,222 15,184,682 60,287,546 75,472,228 7% -8% -5%
South
Heaton
15,267,453 61,412,232 76,679,685 14,228,799 60,075,526 74,304,325 7% 2% 3%
Westgate 22,787,300 72,449,103 95,236,403 27,211,294 47,918,836 75,130,130 -16% 51% 27%
2009 DECC Estimation 2009 Differences2009 Model Estimation
14. • Different performance of the model within the three case study
MLSOAs means interesting insight into the importance of
locality in modelling energy consumption.
• Westgate
– Much more diverse portfolio of building types (e.g. a high
proportion of tall, mixed used buildings);
– Energy systems (e.g. district heating and Economy 7);
− Is a deprived area with a high proportion of social housing;
− Interesting socio-demographic issues at play.
• South Heaton and Castle are much more homogenous in their
structures
15.
16. • In Castle model consistently underestimates gas and electricity.
− Mainly inaccurate assumptions in averaging climate (effects of micro-
climates) and standard occupancy (as residential density increases the
population does not increase at the same rate). Also pensioners and
heating patterns.
• In South Heaton model overestimates the energy consumption in electricity
and gas.
– Mainly combined effect Number of thermal zones, assumptions in
dwelling types enclosed flats, and local building types (Tyneside flats).
There is underestimation on DECC values (ALMO, small commercial
premises categorized as domestic).
• In Westgate, DECC presents underestimation in gas consumption.
– Mainly combined effect of DECC (blocks of flats without individualized
gas meters) and major cohesive energy structures: districts heating
(combined commercial and domestic) and shared properties (with
individual heated rooms).
17. Sample
Property
Type
Property
Age
Floor
Area
(sqm)
Sample
Size
Mean annual
model estimated
heat gas
consumption
(KWh)
Median annual
model estimated
heat gas
consumption
(KWh)
Sample
Size
Mean Annual
DECC
estimated gas
consumption
(KWh)
Median Annual
DECC
estimated gas
consumption
(KWh)
Mean
Difference
(%)
Median
Difference
(%)
Negative
ranks
Positive
ranks
Ties
5 House Bungalow
1965 to
1982
76 to
100
55 16,495 12,557 1,400 13,052 12,557 26% 45% 23 32 0
6 House Bungalow
1945 to
1964
50 to 75 37 12,625 11,764 3,020 10,905 10,326 16% 14% 24 14 0
7 House
Semi-
detached
1945 to
1964
50 to 75 203 15,158 13,118 950 11,070 10,574 37% 24% 117 89 0
8 House
Semi-
detached
1965 to
1982
50 to 75 628 13,951 10,772 290 10,161 9,748 37% 11% 367 282 0
9 House Detached
1965 to
1982
76 to
100
100 18,112 18,599 890 13,796 13,218 31% 41% 27 73 0
10 House Detached
1965 to
1982
101 to
125
18 19,431 20,897 1,410 15,365 14,870 26% 41% 4 14 0
Castle NEED Wilcoxon Signed Ranks
18. • Decrease in heating gas consumption as the floor area
gets smaller for both NEED and NCRM.
19. • wide range of modelled energy consumption within a single property
type and that these are closely linked to specific characteristics of the
properties
20. • uninsulated solid wall
terraced houses. Air-
tightness and
ventilation problems
Low terraces in Byker
21. Tyneside Flats were usually
built to rent. The usable
floor area is below the
average of UK
20,000 Tyneside flats
most having solid walls
the extension is cavity
(hard to treat)
cavity size is too small
(McLoughlin, 2012).
22. Source: Figures before 2003 from DECC Fuel Poverty Monitoring Indicators 2012 (DECC, 2012)
Figures from 2003 onwards from DECC Trends in Fuel Poverty in England 2003 to 2011 (DECC, 2013)
23. • Is ppm use a poor proxy for fuel
poverty?
• Students are not typically
associated with fuel Poverty.
Temporary accommodation.
24. • Castle Bungalows (local area Bungalows) are of bigger floor area
than the North East England (regional Bungalows).
25. • The evidence suggests that energy consumption in domestic homes:
demographic, economic, technical, social and cultural factors and
climate conditions, urban form and building massing.
• Local area characteristics are important factors in energy modelling and
are not always reflected in regional or national indices.
• It can be used to inform and direct policy by testing the effect that
various policy decisions are likely to have on the community energy.
Test the effect of a range of possible future revisions to the Building
Regulations or future revisions in the ECO CSCO (e.g. extend from the
bottom 15% to the bottom 25% most deprived areas, based on the
Index of Multiple Deprivation (IMD)).
• Test potential rules on the identification of hard-to-treat cavities in the
ECO Carbon Emissions Reduction Obligation (CERO).
• Move forward:
− Altering in the energy profile of a group of buildings so that there is
potentially a better match between demand and supply in a
community in a demand side exercise.
• Insights to integrate renewables into the generation portfolio
26. Committee on Climate Change (2013) Reducing the UK’s carbon footprint and managing competitiveness risks. [Online]. Available
at: http://www.theccc.org.uk/wpcontent/uploads/2013/04/CF-C_Summary-Rep_Bookpdf.pdf.
Social Exclusion Unit (2011) A New Commitment to Neighbourhood Renewal - National Strategy Action Plan. [Online]. Available
at:
http://www.bristol.ac.uk/poverty/downloads/keyofficialdocuments/Neighbourhood%20Renewal%20National%20Strategy%20Repor
t.pdf.
Energy Saving Trust, Centre for Sustainable Energy and National Energy Agency (2005) WarmZones external evaluation [Online].
Available at: http://www.cse.org.uk/pdf/warm_zones_evaluation_full_final.pdf.
DECC (2012) Fuel poverty monitoring indicators 2012. [Online]. Available at:
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/66017/5272-fuel-poverty-monitoring-indicators-
2012.pdf.
DECC (2013) Trends in fuel poverty England: 2003 to 2011. [Online]. Available at:
https://www.gov.uk/government/publications/trends-in-fuel-poverty-england-2003-to-2011.