A Stochastic LCA Framework for Embodied Greenhouse Gas Analysis Dr David Shipworth School of Construction Management and Engineering University of Reading.

Slides:



Advertisements
Similar presentations
The delights of data: deficiencies in the quagmire? Angela Druckman and Tim Jackson RESOLVE University of Surrey Carbon Accounting Conference Heriot-Watt.
Advertisements

Canada-U.S. Binational EIO-LCA Model Jonathan Norman Heather L. MacLean Department of Civil Engineering University of Toronto LCA of Oil Sands Technology.
A Tutorial on Learning with Bayesian Networks
Forest Products Industry Calculation Tools Reid Miner, Brad Upton, Jim Palumbo, NCASI Member Associations of ICFPA.
Introduction to Input-Output Based LCA. Admin Issues Friday Feb 16th? 1-2:30 confirmed HERE.
Bayesian estimation in NTA frameworks Details of a data study in progress By Jan W. van Tongeren (ex-Head National Accounts, UNSD) and Arjan Bruil (statistical.
Analysis of CO 2 Abatement Strategies in China’s Electricity Sector Hu Junfeng ( 胡军峰 ) North China Electric Power University July, 2010.
Embodied Carbon: The Inventory of Carbon & Energy Database 28 th September 2011 Dr. Craig Jones
Carbon Management and Infrastructure Development: Lessons from Industrial Ecology Dr. Robin Curry Institute for a Sustainable World Queen’s University,
Economic Input-Output Life Cycle Assessment / Life Cycle Assessment and Green Design.
The UK Climate Change Levy and Ecological Tax Reform Professor Stephen Smith Department of Economics University College London.
Goal: Reconstruct Cellular Networks Biocarta. Conditions Genes.
End of Chapter 8 Neil Weisenfeld March 28, 2005.
SGM P.R. Shukla. Second Generation Model Top-Down Economic Models  Project baseline carbon emissions over time for a country or group of countries 
Climate change AEBN Caltex speech ppt Australian Environment Business Network Industry and Greenhouse Conference Sydney, 27 February 2008 Alternative.
>> Input-output tables, models and environmental- economic analysis Ole Gravgård SEEA Training Seminar for the ECA Addis Ababa 2-5 February 2015.
Lecture II-2: Probability Review
5 th Meeting of the UNCEEA June 23-25, New York Pablo Munoz and Brad Ewing Harmonizing the National Footprint Accounts with the System of Integrated Environmental.
1 D r a f t Life Cycle Assessment A product-oriented method for sustainability analysis UNEP LCA Training Kit Module k – Uncertainty in LCA.
Super-Resolution of Remotely-Sensed Images Using a Learning-Based Approach Isabelle Bégin and Frank P. Ferrie Abstract Super-resolution addresses the problem.
Biomass Carbon Neutrality in the Context of Forest-based Fuels and Products Al Lucier, NCASI Reid Miner, NCASI
Laboratory of ecosystem management Life Cycle Environmental Impacts of Internet Yves Loerincik, Olivier Jolliet, Greg Norris.
African Centre for Statistics United Nations Economic Commission for Africa Chapter 11: Chapter 11: Supply and use tables to input-output tables Ramesh.
1 Instruments and statistics - The response ® in the DPSIR Torstein Bye 1 and Annegrete Bruvoll 2 1 Director Economics, Energy and Environment, Statistics.
Global Warming Effect Assessment in the Electricity Sector Using Hybrid Life-cycle Inventory Assessment Arpad Horvath, Assistant Professor Department of.
1 NATIONAL STATISTICAL COORDINATION BOARD International Workshop on Energy Statistics, Beijing, China, September 2012 Vivian R. Ilarina, Philippines.
NRG 173: Carbon Footprints for Climate Action in Complex Organizations Spring Term 2011 Class 11 of 20 May 5, 2011 Kelly Hoell Good Company Eugene, OR.
Sustainability at NMSU A part of the integrated planning process.
PRICE AND VOLUME MEASURES NATIONAL ACCOUNTS STATISTICS WORKSHOP PRICE AND VOLUME MEASURES Workshop on national accounts for Asian member countries of the.
LIEN – Reporting Energy & CO 2 Emissions for Carbon Tax and Emissions Trading. Fuel Use and Distribution in Ireland.
Overview of Economic Methods to Simulate Land Competition Forestry and Agriculture Greenhouse Gas Modeling Forum National Conservation Training Center.
Compilation of emission inventories The situation in the Netherlands Special Session of the UNCEEA on Climate Change (New York, 25 June afternoon)
Methods and Techniques of Planning Methods and Techniques of Planning Krishna Khadka 1 krishna Khadka.
Ucertainty estimates as part of the inventory process Kristin Rypdal, CICERO.
1 “Using Carbon Markets to Encourage the Uptake of Low Carbon Vehicles” Meeting the Low Carbon Challenge The Low Carbon Vehicle Partnership Third Annual.
Monte Carlo Methods Versatile methods for analyzing the behavior of some activity, plan or process that involves uncertainty.
1Jukka Muukkonen Carbon binding and forest asset accounts Forest related issues in greenhouse gas inventory Connections between SEEA2003 forest asset accounts.
The Dutch energy accounts Sjoerd Schenau Statistics Netherlands.
The Rights and Wrongs of Carbon Accounting About life cycles and footprints in the carbon world Carbon Accounting Conference Heriot-Watt University, Edinburgh.
Mixture Models, Monte Carlo, Bayesian Updating and Dynamic Models Mike West Computing Science and Statistics, Vol. 24, pp , 1993.
Input-Output Analysis of Philippine Sectoral CO 2 Emissions Raymond R. Tan, Ph.D. 1, Joel Q. Tanchuco, M.A. 2 1 Center for Engineering and Sustainable.
Problem Limited number of experimental replications. Postgenomic data intrinsically noisy. Poor network reconstruction.
The Measurement of Direct Value of Forage Production Ken Belcher Department of Bioresource Policy, Business and Economics November 20, 2008.
1 Hybrid LCA Models. 2 Admin Issues Next 2 weeks: –Heather, me, others out of town –3 guest lectures (hybrid methods and uncertainty) Last week of class.
Setting the Boundaries Stephen Boyle, Principal Consultant, Climate Change and Policy 11 th March 2009 A world leading energy and climate change consultancy.
Sustainable growth with renewable and fossil fuels energy sources Carlo Andrea Bollino, Silvia Micheli 30 th USAEE/IAEE North American Conference October.
The generalization of Bayes for continuous densities is that we have some density f(y|  ) where y and  are vectors of data and parameters with  being.
Markov Chain Monte Carlo for LDA C. Andrieu, N. D. Freitas, and A. Doucet, An Introduction to MCMC for Machine Learning, R. M. Neal, Probabilistic.
1 1 SEEA/Environmental accountings’ user needs with regards to energy statistics Julie L. Hass, Ph.D. Kristine E. Kolshus Division for Environmental Statistics.
Simon Ramsay President. Rail Freight Grains industry dependent on efficient rail transport Average of 2 million tonnes of grain moves by rail Predicted.
CO 2 emissions on a quarterly basis Maarten van Rossum and Sjoerd Schenau Presented by Ellen Brinksma.
1 Towards the implementation of SEEA-E Learning Centre on Environmental Accounts UNSC, New York February 2009 Alessandra Alfieri United Nations Statistics.
Bayesian Optimization Algorithm, Decision Graphs, and Occam’s Razor Martin Pelikan, David E. Goldberg, and Kumara Sastry IlliGAL Report No May.
S T A T I S T I K A U S T R I A Energy Balances, Energy Accounts and the world beyond Wolfgang Bittermann 12th Meeting of the London City.
Scottish National Accounts Project (SNAP) Andrew Mortimer SNAP team.
1Jukka Muukkonen Carbon binding of forests: some remarks on classification and valuation 13 th London Group Meeting
1 Life Cycle Assessment A product-oriented method for sustainability analysis UNEP LCA Training Kit Module k – Uncertainty in LCA.
Canadian Bioinformatics Workshops
US I/O Model(s) and the integration with other data sets Gregory A. Norris Sylvatica / Harvard University / U. New Hampshire USA.
Measuring the Health impacts of NHS travel
School of Earth & Environment
CS 367 – Introduction to Data Structures
Environmental input-output analysis at Statistics Netherlands
SUTs – data sources and bridge tables
The importance of energy balances
Eurostat – Unit E2 Frédéric NAUROY
Introduction to the System of Environmental-Economic Accounts Central Framework (SEEA-CF 2012) European Statistical Training Programme (ESTP): Environmental.
SUTs – data sources and bridge tables
Input-output tables THE CONTRACTOR IS ACTING UNDER A FRAMEWORK CONTRACT CONCLUDED WITH THE COMMISSION.
Input-output tables Robin Lynch
Presentation transcript:

A Stochastic LCA Framework for Embodied Greenhouse Gas Analysis Dr David Shipworth School of Construction Management and Engineering University of Reading - UK

Objectives Model effect of policies encouraging low carbon technologies (e.g. Carbon taxes & Emission Trading) Model effect of policies encouraging low carbon technologies (e.g. Carbon taxes & Emission Trading) Avoid misrepresentation of single ‘average’ CO 2eq figures for materials Avoid misrepresentation of single ‘average’ CO 2eq figures for materials To capture lost information (variance, skewness, etc) To capture lost information (variance, skewness, etc) Model the ‘carbon diversity’ of materials Model the ‘carbon diversity’ of materials

Requirements of model Stochastic: Stochastic: Require probability distributions of embodied CO 2eq in materials Require probability distributions of embodied CO 2eq in materials Complete: Complete: Incorporate the system boundary completeness of Input-Output (IO) with the product specificity of Process Analysis (PA) (a ‘hybrid’ model) Incorporate the system boundary completeness of Input-Output (IO) with the product specificity of Process Analysis (PA) (a ‘hybrid’ model) Evolutionary: Evolutionary: Model to support progressive integration PA data as and when it becomes available Model to support progressive integration PA data as and when it becomes available

Data Sources UK National Environmental Accounts (UKNEA) UK National Environmental Accounts (UKNEA) 91 sector IO accounts 91 sector IO accounts Aggregated for environmental homogeneity Aggregated for environmental homogeneity UK National Atmospheric Emissions Inventory (NAEI) UK National Atmospheric Emissions Inventory (NAEI) ~4400 emissions estimates by economic sector, source and fuel (thousand tonnes) for C, CH 4 & N 2 O ~4400 emissions estimates by economic sector, source and fuel (thousand tonnes) for C, CH 4 & N 2 O Includes non-fuel emission sources Includes non-fuel emission sources UK Annual Business Inquiry (ABI) UK Annual Business Inquiry (ABI) Total purchases data for 3-digit sub-sectors at basic prices Total purchases data for 3-digit sub-sectors at basic prices Existing process analysis data Existing process analysis data Anonymous, process level data by UKNEA sector Anonymous, process level data by UKNEA sector

Components of model: Expanded UKNEA UKNEA is 91x91 Environmental I-O matrix UKNEA is 91x91 Environmental I-O matrix Transaction between sectors is in £M Transaction between sectors is in £M Annual emissions vectors allow conversion to emissions flows (T.CO 2eq ) or intensities (T.CO 2eq /£M) Annual emissions vectors allow conversion to emissions flows (T.CO 2eq ) or intensities (T.CO 2eq /£M) Each sector contains between 0 and 9 SIC 3-digit sub-sectors Each sector contains between 0 and 9 SIC 3-digit sub-sectors Expanding to sub-sectors creates 91 by 161 (2-digit by 3-digit) matrix Expanding to sub-sectors creates 91 by 161 (2-digit by 3-digit) matrix

Components of model: Expanded UKNEA New column totals available from ABI New column totals available from ABI New 3-digit sub-sector row transaction totals are existing 2-digit transaction values New 3-digit sub-sector row transaction totals are existing 2-digit transaction values 2-digit sales to 3-digit sub-sectors reconstructed using GME method 2-digit sales to 3-digit sub-sectors reconstructed using GME method Product can be viewed either as: Product can be viewed either as: a 91 by 161environmental IO table; or a 91 by 161environmental IO table; or a 91x91 IO table with cells containing multi-state data a 91x91 IO table with cells containing multi-state data

Components of model: Emissions Intensities Use ~4400 NAEI data for C, CH 4 & N 2 O Use ~4400 NAEI data for C, CH 4 & N 2 O Allocate to SIC 3-digit (161) sub-sectors based on primary sector definitions Allocate to SIC 3-digit (161) sub-sectors based on primary sector definitions Gives total emissions from 3-digit sub-sector Gives total emissions from 3-digit sub-sector Use ABI data to convert to emissions intensities (T/£M) at the 3-digit level Use ABI data to convert to emissions intensities (T/£M) at the 3-digit level

Components of model: Bayesian Prior Apply 3-digit sub-sector emissions intensities to reconstructed transaction values between 2-digit sectors and 3-digit sectors Apply 3-digit sub-sector emissions intensities to reconstructed transaction values between 2-digit sectors and 3-digit sectors This gives Dirichlet emissions intensity distribution within each 2-digit sector This gives Dirichlet emissions intensity distribution within each 2-digit sector The number of states of the Dirichlet distribution equals number of 3-digit sub- sectors The number of states of the Dirichlet distribution equals number of 3-digit sub- sectors

Components of model: Process Analysis Data Use anonymous process level data Use anonymous process level data Data collected by UK ETS sector level entrants Data collected by UK ETS sector level entrants System boundary is UKNEA sector definition System boundary is UKNEA sector definition Data expressed as T.CO 2eq /£M Data expressed as T.CO 2eq /£M Represent data as multinomial distribution Represent data as multinomial distribution

Components of model: Bayesian integration Integrate prior I-O distribution (Dirichlet), with process data process distribution (Multinomial) Integrate prior I-O distribution (Dirichlet), with process data process distribution (Multinomial) Done using Markov Chain Monte Carlo package (WinBUGS) Done using Markov Chain Monte Carlo package (WinBUGS) Resulting ‘Posterior’ distribution is most heavily influenced by the stronger data set Resulting ‘Posterior’ distribution is most heavily influenced by the stronger data set New data can continually be integrated New data can continually be integrated

The UKNEA in graph theory terms The UKNEA is a 91 sector (node) deterministic graph The UKNEA is a 91 sector (node) deterministic graph Connected sectors are linked by a single pathway (edge) Connected sectors are linked by a single pathway (edge) The I-O matrix is the ‘adjacency’ matrix of this graph – a value in a cell indicates a pathway between sectors The I-O matrix is the ‘adjacency’ matrix of this graph – a value in a cell indicates a pathway between sectors Each pathway has an emissions intensity Each pathway has an emissions intensity Total emissions into a sector are found by tracing back along the carbon pathways Total emissions into a sector are found by tracing back along the carbon pathways The pathways create a carbon ‘tree’ for that sector The pathways create a carbon ‘tree’ for that sector

The Model in graph theory terms The Model is a 91 sector stochastic graph The Model is a 91 sector stochastic graph Connected sectors are linked by one or more pathways Connected sectors are linked by one or more pathways The expanded I-O matrix is the ‘adjacency’ matrix of this graph – a distribution in a cell indicates multiple pathways between sectors The expanded I-O matrix is the ‘adjacency’ matrix of this graph – a distribution in a cell indicates multiple pathways between sectors The distributions combine prior I-0 data integrated with process analysis data on an ongoing basis The distributions combine prior I-0 data integrated with process analysis data on an ongoing basis

The Model in graph theory terms Each of the pathways has a different emissions intensity Each of the pathways has a different emissions intensity Total emissions into a sector are found by tracing back along the carbon pathways – where there are multiple pathways one is chosen at random Total emissions into a sector are found by tracing back along the carbon pathways – where there are multiple pathways one is chosen at random The set of all possible pathways creates a carbon ‘forest’ for that sector The set of all possible pathways creates a carbon ‘forest’ for that sector The carbon diversity of the forest is the carbon diversity of the sector The carbon diversity of the forest is the carbon diversity of the sector