Deconstruct: A scalable method of as-built heat power loss coefficient inference for UK dwellings using smart meter data

TLDR: DecDeconstruct as mentioned in this paper is a method of estimating the as-built Heat Power Loss Coefficient (HPLC) of occupied dwellings as a measure of thermal performance, using just smart-meter and meteorological data.

About: This article is published in Energy and Buildings.The article was published on 2019-01-15 and is currently open access. It has received 19 citations till now. The article focuses on the topics: Smart meter.

Figures

Table 1: Summary of model parameters and source of parameter value. 208

Table 2: Descriptive statistics of the HPLC estimates for the EDRP-EDF data (inferred using 358 Deconstruct) and the CHM data (calculated bottom-up). 359

Table 3: Description of parameter uncertainties used as inputs for the Monte Carlo uncertainty 389 estimate. 390

Frequently asked questions

Q1. What are the contributions in this paper?

Chambers et al. this paper developed a method to characterise `` as-built '' thermal performance of individual UK dwellings with respect to heating, independently of occupant thermal behaviours, that can be performed rapidly and nonintrusively at scale, using smart meter data collected from large numbers of dwellings. 

Q2. How many buildings were used to calculate the national average?

EFUS provided 323 per-site weightings which can be used to calculate nationally representative statistics using the 324 buildings sampled. 

Q3. What is the need for a special consideration for the dwelling energy balance?

Since the energy use of these heating systems is metered there is no need for special 496 consideration for the dwelling energy balance. 

Q4. What is the way to infer dwelling parameters from the 174 simplified thermal model?

During this collection period, natural variability in weather and power demand will result in a 173 subset of days during which conditions are optimal for inferring dwelling parameters using the 174 simplified thermal model. 

Q5. How can the authors control the heating system?

shorter sampling periods could be achieved through a more pro-active data 446 collection strategy, for example by sampling specific periods during winter and summer or controlling 447 the heating system in a structured way (for example using a smart heating system to perform a 448 controlled test when occupants are absent). 

Q6. What is the way to infer dwelling parameters from metered energy demand data?

In order to infer dwelling parameters from metered energy demand data, 169 a ‘post-hoc control trial’ methodology was developed, which makes use of a structured sampling of 170 accumulated smart meter data to produce robust parameter estimates. 

Q7. How can a deconstruct be applied to large numbers of dwellings?

64The method aims to be scalable in that it may readily be applied to large numbers of dwellings without 65 incurring significant manual effort, costs, or being computationally prohibitive. 

Q8. What is the way to estimate the thermal performance of a dwelling?

The Deconstruct method is ideally placed to take 524 advantage of this data source to provide thermal performance estimates for dwellings and provide 525 additional value from meter data for utilities and occupants. 

Q9. How robust was the HPLC 450 to weather changes?

The HPLC 450 was found to depend on the internal temperature model approximation for 𝐹 , but was robust to 451 changes in 𝜂 , 𝑃 and 𝑇 . 

Q10. What is the power demand for the highest temperature data point in the sample?

228 The power demand 𝑃 for the highest temperature data point in the sample was used as a lower 229 cut-off threshold for the power value, removing all points where 𝑃 < 𝑃 (Figure 2). 

Q11. What could be the useful information about the occupancy heuristic?

Smart thermostats could provide valuable additional 487 information in this regard, as they are usually designed to adapt heating patterns to occupancy and 488 do so by using a range of methods to predict occupancy, such as drawing on data from smartphone 489 apps. 

Q12. What was the geographical identifier used to link dwelling monitoring data with external 305 temperature and?

The Government Office 304 Region (GOR) geographical identifier was used to link dwelling monitoring data with external 305 temperature and solar irradiance from the MetOffice.