A Tool for Energy Planning

and GHG Mitigation

Assessment
Seksan Udomsri - KTH
Charles Heaps - SEI
Stockholm Environment Institute

An Independent Research Affiliate of Tufts University

Charles Heaps, Ph.D.

Stockholm Environment Institute
U.S. Center
11 Curtis Avenue
Somerville, MA 02144, USA
Tel: +1 (617) 627-3786
Fax: +1 (617) 449-9603

Web: www.sei-us.org
Email: info@sei-us.org
2

Stockholm Environment Institute

An independent international research organization focusing on the issue of

sustainable development.
Headquarters in Stockholm, Sweden with centers in the US, UK (York &
Oxford), Estonia, and Bangkok.
Main program areas: climate & energy, water resources & ecological
sanitation, atmospheric pollution, risk, livelihoods & vulnerability,
sustainable futures.
Apx. 150 staff (20 in the U.S.).
Funders include the Swedish Government, multilateral agencies,
foundations and national & local governments.
SEI-US is an independent non-profit research institute affiliated with Tufts
University in Massachusetts.
Web sites: www.sei-us.org and www.sei.se

Todays Talk
Part 1: Energy Planning, GHG Mitigation
Assessment and Energy Modeling
Part 2: LEAP Overview
Part 3: LEAP Demonstration

Part 1: Energy Planning, GHG

Mitigation Assessment and
Energy Modeling

Why Energy Planning is

Important

General goal: matching supply to demand at reasonable cost.

Energy is an area of the economy where a long-term perspective
and active planning and policy-making are vital.

A major driver of emissions and climate change.

A major cause of other environmental impacts
A major economic cost (and vulnerability) and a vital basic need.
Tendency toward natural monopoly for delivery of some energy forms
(electricity power lines, gas pipes, etc.) and often significant market
power of major energy companies.
Long life of energy equipment (cars have ~15 year life; power plants ~
50 years; housing ~100 years; urban development has implications for
centuries).

Energy planning can therefore have potentially a huge impact on

societies.
Forecasting with any certainty has proven very difficult.
Traditional energy policy analyses (e.g. least cost optimal
planning) may not be well suited to the coming climate challenge,
where social choice may be as important as technical fix, and where
robust planning rather than optimal solutions are needed.
6

Cost-benefit analysis: pros and cons

Cost-benefit analysis is now seen as the standard economic

method for policy analysis
Add up all costs of a policy, and all benefits
Approve the policy if benefits are greater

Cost-benefit analysis is a powerful tool but not the right tool for
every job

Strong simplifying assumptions make it powerful but also limit its

applicability
All costs and benefits must have prices
Total costs to society are compared to total benefits, regardless of
who pays or who gains
When future costs or benefits are uncertain, an average or most likely
value is used

Worst case or average?

Economic analysis often relies on average forecasts
Sea level rise: without catastrophic loss of ice sheets, less
than 1 meter forecast for this century
Will be hard on low-lying areas (Bangladesh, Miami, Venice)

The greatest fears about climate change are often

based on worst-case possibilities
Complete loss of the Greenland (or West Antarctic) ice sheet
would cause 7 meters of sea level rise
Catastrophic impacts on most coastal cities, communities

Will the Greenland ice sheet melt?

Complete melting is still unlikely

But it becomes less unlikely as temperatures rise
Average: no problem this century
Worst case: increasing cause for worry

Things that wont happen (soon)

Why buy insurance?

People care a lot about unlikely worst cases
How much time do you leave to get to the airport?
Airport security is all about worst case possibilities

Insurance is not based on average outcomes

The average (US) house has a fire every 250 years (0.4%
probability per year of a residential fire)
But most people have fire insurance

Probability of death next year is less than 1% until age 61; under
0.2% until 40 (US data)
But most young parents have life insurance

Probability of enough warming to guarantee loss of

Greenland ice sheet is much greater than 1%
Should we buy insurance for the planet?

On average, sea walls are not needed

 Uncertainty and catastrophic risk are decisive

Climate policy is insurance against low-probability (but not
impossible) catastrophic events
By comparison, the most likely outcome is irrelevant

Climate catastrophes are now at least as likely as risks

(fire, death) we buy insurance against
Exact probabilities are unknown, but become more likely as
the climate changes

Cost-benefit analysis offers to guard against the risk of

spending too much on renewable energy, etc.
This is a very different (less urgent) problem

The real economic question: what is the least-cost way

to ensure that we prevent global catastrophe?

IRP vs. Traditional Utility

Planning
Traditional energy planning
Focus on demand growth projections
Expansion planning to determine available resources and when
they are needed
Production cost analysis to rank supply options by cost
Calculation of required revenues and rates

Integrated Resource Planning (IRP)

Meet demand for energy services instead of energy
Focus on Demand Side Management (DSM) and efficiency
programs

Include externalites in decision making

Emissions costs
Social costs

13

Traditional Utility Planning

Process:
Expand supply resources to meet anticipated energy demand growth.

Objectives:
high reliability (wide reserve margins)
least cost expansion planning

Results:
Rapid capacity expansion
Promotion of demand growth
Little consideration of the necessity for energy efficiency

14

Integrated Resource
Planning
Process:
Integrated assessment of supply and demand-side options in order to
meet the projected demand for energy services.

Objectives:
Least total cost (economic + social + environmental)

15

Forecasts can be Wrong

Successive 10 year forecasts of U.S. Summer peak electricity demand issued by

the North American Electric Reliability Council (1974-1983)
16
Source: Smil, 2003

Forecasts or Scenarios?
Energy Forecasts
What is likely?

Energy Scenarios

Approach

Rational focus on analysis and

outcomes

What could be?

Under what assumptions?
Focus on process. strategy and
learning

Objective

To develop the most likely pathway

and characterize uncertainty

To develop a number of insightful

pathways that explore uncertainties

Methods

Analytical models and driver variables

Qualitative stories, quantified and

evaluated by models

Treatments of uncertainty Probabilistic methods, statistics and

transparency of assumptions

Exploration of critical uncertainties,

and separation of predetermined and
uncertain elements in crafting stories

Important Actors

Group facilitators, strategists,

problem-solvers

Reliance on experts, state and

national planning agencies

17

Forecasting & Backcasting

Where is society going?

forecast

?
Where do we want to go?
How do we get there?
backcast
18

Mitigation Assessments
A specific application of Integrated Planning.
Designed to:
To provide policy makers with an evaluation of technologies and
practices that can mitigate climate change and also contribute to
national development objectives.
Help us to understand the costs of avoiding climate disruption.
Identify potential project/programme investments.

Steps in a Mitigation Assessment

Define Time Frame (typically long run)

Define Scope (energy demand & supply, agriculture, land-use,
forestry, solid waste, geological sequestration).
Define participants and key stakeholders (policy makers, scientific
community, NGOs).
Define desired results.
Select methodologies consistent with data and expertise availability.
Standardize key parameters (base year, end year, discount rate,
etc.)
Define project boundaries (consistent with approach used to
develop emissions inventories)
Define scenarios (typically at least two: baseline and mitigation)

Structure of a Mitigation
Assessment

Source: UNEP Economics of Greenhouse Gas Limitations Guidelines (1999)

Steps of a Mitigation
Assessment
Depends on goals, scope & sectors, but has common steps:
1. Collect data.
2. Assemble base year/historical data on activities,
technologies, practices and emission factors.
3. Calibrate base year to standardized statistics such as
national energy balance or emissions inventory.
4. Prepare baseline scenario(s).
5. Screen mitigation options.
6. Prepare mitigation scenario(s) and sensitivity analyses.
7. Assess impacts (social, economic, environmental).
8. Develop Mitigation Strategy.
9. Prepare reports.

Timeframe for Assessments

Ideally, should be long-term to reflect economic lifetime
and potential for stock turnover of major technologies
(e.g. 30-40 years in the energy sector).
But development of long-term projections are very
difficult, especially in developing countries, due to
uncertainties over future development and limited
statistical data.
Nearer term assessments (10-20 years) based on
national plans and sectoral assessments are more
practical for most developing countries.
These nearer term assessments could usefully be
complemented by more aggregate assessments of
longer-term trends.

Key Study Parameters

Base year of study

Time horizon
System boundaries
Costing perspective (societal or market)
Discount rate
Treatment of avoided emissions:
Should they be discounted?

Discounting Over Long

Timeframes

Discounting states that money (excluding inflation) is worth more

today than tomorrow

There is criticism of discounting because it devalues the

future worth of things
100%
3%
10%

90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0

20

40

60
Future Year

80

100

25

Sustainability and Discounting

Sustainability - "Meeting the needs of the present generation without

compromising the ability of future generations to meet their needs. Brundtland Report
Intergenerational equity and sustainability would seem to indicate
the use of a 0% discount rate
Costs and benefits that occur today or in the future should not be valued
differently.

Discounting leads to outcomes that may not be equitable

Future costs will be lower than present costs and distant future costs
are close to zero in present terms.
Thus, future costs (e.g. climate change damage, impacts of nuclear
waste) may be undervalued compared with avoiding present costs.
However, substitution of capital implies that money saved today w.ill
grow and can be used to mitigate or adapt to future damages

26

Data Collection
Specific data requirements depend on scope and
objectives of study.
Depending on methodology, may need to collect data
only for a base year, but longer historical time series
data generally provides a better context and may be
required for econometric analyses.
Decide on level of data disaggregation: avoid temptation
to be data driven.
Primary focus should be collation of secondary data, but
some primary data collection may be required and
assumptions/judgment will be needed to fill data gaps.

Key Participants

The development of mitigation assessments will require close cooperation

among a wide range of stakeholders.
Energy, environment and finance ministries will all likely need to be
involved. Some tasks may be undertaken by outside consultants or the
academic community.
Expert skills required include: statisticians, energy policy experts, engineers,
modelers, statisticians & technical writers.
However, mitigation assessments are not simply technocratic exercises:
they involve much broader judgments about how mitigation activities can fit
into national development priorities.
Thus, the context for defining mitigation priorities will in large part depend
on the process by which priorities are expressed in each country (e.g.
whether priorities are set by the Government alone or in consultation with
other stakeholders such as NGOs, industries, the scientific community, etc.)

Steps of a Mitigation
Assessment
Depends on goals, scope & sectors, but has common steps:
1. Collect data.
2. Assemble base year/historical data on activities,
technologies, practices and emission factors.
3. Calibrate base year to standardized statistics such as
national energy balance or emissions inventory.
4. Prepare baseline scenario(s).
5. Screen mitigation options.
6. Prepare mitigation scenario(s) and sensitivity analyses.
7. Assess impacts (social, economic, environmental).
8. Develop Mitigation Strategy.
9. Prepare reports.

Baseline Scenarios

Plausible and consistent description of how a system might evolve into the future
in the absence of explicit new GHG mitigation policies.
Assessments will typically require one or more baseline scenarios: the
counterfactuals against which mitigation measures will be evaluated.
Critical to a mitigation assessment since mitigation measures are largely judged
on the basis of the incremental costs and benefits relative to the baseline.
Should not be considered a forecast of what will happen in the future, since the
future is inherently unpredictable and depends, in part, on planning and the
adoption of policies.
Highly uncertain over the long run and may be controversial. For example,
should a baseline assume that the Millennium Development Goals will actually be
met, and if so what does this imply for the energy systems of the poorest
countries?
Ideally, multiple baselines should be constructed to reflect uncertainties
(sensitivity analysis). Each baseline requires separate mitigation analyses.
Baselines should not be simple extrapolations of current trends: they should
consider likely evolution of activities that effect emissions and sinks including:
Macroeconomic and demographic trends.
Structural shifts in the economy
Evolution of technologies and practices, (saturation effects, likely adoption of
efficient technologies).

Steps of a Mitigation
Assessment
Depends on goals, scope & sectors, but has common steps:
1. Collect data.
2. Assemble base year/historical data on activities,
technologies, practices and emission factors.
3. Calibrate base year to standardized statistics such as
national energy balance or emissions inventory.
4. Prepare baseline scenario(s).
5. Screen mitigation options.
6. Prepare mitigation scenario(s) and sensitivity analyses.
7. Assess impacts (social, economic, environmental).
8. Develop Mitigation Strategy.
9. Prepare reports.

Screening Mitigation Options

Enables a rough assessment of the potential feasibility of options.

Particularly important when using bottom-up methodologies in which a wide
range of technologies and policies need to be considered.
May include a quantitative assessment of the mitigation potential (T CO2)
and cost of saved carbon ($/TC) of each option. May also include qualitative
factors.
One approach is to prepare a matrix and assign scores or rankings to
options in order to identify those options that need to be included in the
more in depth analysis.
Gives the opportunity to explicitly consider a comprehensive set of options
while reducing the level of effort required in the later more in-depth
mitigation analysis.
Reduces likelihood of overlooking important options.
Screening criteria should be consistent with overall framing of mitigation
scenario.

Possible Screening Criteria

Potential for large impact on greenhouse gases (GHGs)

Consistency with national development goals
Consistency with national environmental goals, such as:

emissions reduction of local air pollutants

effect on biodiversity
soil conservation
watershed management
indoor air quality, etc.

Potential effectiveness of implementation policies

Sustainability of an option
Data availability for evaluation
Institutional considerations such as:
Institutional capacity needed (data collection, monitoring, enforcement,
permitting, etc.)
Political Feasibility
Replicability (adaptability to different geographical, socio-economic-cultural,
legal, and regulatory settings)

Other sector-specific criteria

Screening Matrix
Examples of Criteria
Mitigation Potential
Direct Costs
Indirect Costs
- Increase in domestic employment
- Decrease in import payments
Consistency with Development Goals
- Potential for wealth generation
- Consistency with MDGs
Consistency with Environmental Goals
-Potential for reducing air, water and other pollution
Long term sustainability of option
Data
-Availability
-Quality
Feasibility (political, social, technical)

Mitigation Option 1
Tonnes CO2,
score or ranking (low, medium)
$/Tonne, C/B ratio, score or ranking
Score or ranking
Score or ranking
Score or ranking
Score or ranking
Score or ranking
Score or ranking
Score or ranking
Score or ranking
Score or ranking

Option 2

Option 3

Screening with Cost Curves

A technique for screening

and ranking GHG mitigation
options.
Plot GHG reduction from
successive mitigation options
(e.g. tonnes of CO2 avoided)
against cost per unit of GHG
reduction (e.g. $/tonne).
Area under curve yields total
cost of avoided emissions.
Need to consider
interdependencies among

Source: Sathaye & Meyers. Greenhouse Gas Mitigation Assessment: A Guidebook (1995)

Three Approaches to
Developing Cost Curves
Partial approach
Retrospective systems approach
Integrated approach

The Partial Approach

Each technology is evaluated separately
and compared to some reference
technology.
Overall emissions reductions and costs
are created by combining options while
assuming no interaction between options.

The Retrospective System

Approach
Step 1: Independent ranking of options (like partial
approach)
Step 2: Include most cost effective option in a scenario
and then recalculate costs and emission reductions for
all other options.
Step 3: Include next option and recalculate.
Continue until cost curve meets mitigation objectives.
Takes into account interdependencies between each
option and previous options on the curve.
May not account for impacts that more expensive
options have on cheaper options already chosen.

Integrated Approach
Requires an integrated model that can chose marginal
options based on their cost per emission reduction.
Automatically develops least cost curves within technical
parameters and model constraints.
Fully accounts for interdependencies among options.
Powerful but complex modeling process also may be
difficult to equate reductions with specific options (i.e.
points on cost curve are due to some interaction of
options).
This is important when considering who pays for an option (or who gets
the rewards from a CDM investment).
Need to trade-off accuracy vs. complexity. The complexity of assessing
options when dealing with theoretical counterfactuals makes for high

Steps of a Mitigation
Assessment
Depends on goals, scope & sectors, but has common steps:
1. Collect data.
2. Assemble base year/historical data on activities,
technologies, practices and emission factors.
3. Calibrate base year to standardized statistics such as
national energy balance or emissions inventory.
4. Prepare baseline scenario(s).
5. Screen mitigation options.
6. Prepare mitigation scenario(s) and sensitivity analyses.
7. Assess impacts (social, economic, environmental).
8. Develop Mitigation Strategy.
9. Prepare reports.

Mitigation Scenarios
Reflect a future in which explicit policies and measures
are adopted to reduce the sources (or enhance the
sinks) of GHGs.
Mitigation scenarios should take into account the
common but differentiated responsibilities of the Parties
and the specific national and regional development
priorities, objectives and circumstances.
Mitigation scenarios should not simply reflect current
plans. Instead they should assess what would be
hypothetically achievable based on the goals of the
scenario.

Steps in Constructing
Mitigation Scenarios
Establish framing.
Create option portfolios (identify synergistic and/or mutually
exclusive options & double counting), estimate penetration
rates.
Construct integrated scenarios using chosen modeling
methodology.
Calculate overall costs, benefits and GHG mitigation potential.

Mitigation Scenarios
Scenario frameworks may include:
An emission reduction target
relative to the baseline,
relative to emissions in some historical year, or
Relative to some indicator such as CO2/capita or CO2/$

All options up to a certain cost per unit of emissions

reduction (equivalent to a carbon tax).
No regrets (cost-effective options only).
Specific options or technologies: included based on
perceived technical and/or political feasibility.

Part 2: Modeling Methods and

a LEAP Overview

Use of Models in Mitigation

Assessments: Why Use a
Model?
Reflects complex systems in an
understandable form.
Helps to organize large amounts of data.
Provides a consistent framework for
testing hypotheses.

45

Top-Down Models

Examine general impact on economy of energy policies.

Typically examine variables such as GDP, employment, imports,
exports, public finances, etc.
Assume competitive equilibrium and rational behavior in consumers
and producers.
Tend to be country-specific. Off-the-shelf software not typically
available.
Can be used in conjunction with bottom-up approaches to help
check consistency.
E.g. energy sector investment requirements from a bottom-up
energy model used in macroeconomic assessment to check the
GDP forecasts driving the energy model.

Energy Sector Assessment

Models
Bottom-up

Top-down

Use aggregated economic data

Assess costs/benefits through

impact on output, income, GDP
Implicitly capture administrative,
implementation and other costs.
Assume efficient markets.

Capture intersectoral feedbacks

and interactions
Commonly used to assess impact
of carbon taxes and fiscal policies
Not well suited for examining
technology-specific policies.

Use detailed data on fuels,

technologies and policies
Assess costs/benefits of individual
technologies and policies
Can explicitly include administration
and program costs
Dont assume efficient markets,
overcoming market barriers can
offer cost-effective energy savings
Capture interactions among
projects and policies
Commonly used to assess costs
and benefits of projects and
programs

Bottom-up Energy Policy Models

Optimization Models
Typically used to identify least-cost configurations of energy systems
based on various constraints (e.g. a CO2 emissions target)
Selects among technologies based on their relative costs.

Simulation Models
Simulate behavior of consumers and producers under various signals
(e.g. prices, incomes, policies). May not be optimal behavior.
Typically uses iterative approach to find market clearing demandsupply equilibrium.
Energy prices are endogenous.

Accounting Frameworks
Rather than simulate the behavior of a system in which outcomes are
unknown, instead asks user to explicitly specify outcomes.
Main function of these tools is to manage data and results.

Hybrids Models combining elements of each approach.

48

Optimization Models (1)

Typically uses linear programming to identify energy systems that

provide the least cost means of providing an exogenously
specified demand for energy services.
Optimization is performed under constraints (e.g. technology
availability, supply = demand, emissions, etc.)
Model chooses between technologies based on their lifecycle
costs.
Least-cost solution also yields estimates of energy prices (the
dual solution).

49

Optimization Models (2)

Pros:
Powerful & consistent approach for a common type of analysis called
Backcasting. E.g. What will be the costs of meeting a certain policy goal?
Especially useful where many options exist. E.g. : What is the least cost
combination of efficiency, fuel switching, pollution trading, scrubbers and low
sulfur coal for meeting a SOx emissions cap?

Cons:
Questionable fundamental assumption of perfect competition (e.g., no
monopolistic practices, no market power, no subsidies, all markets in
equilibrium).
Not well suited to simulating how systems behave in the real world.
Assumes energy is only factor in technology choice. Is a Ferrari the same as a
Ford? Tends to yield extreme allocations, unless carefully constrained.
Not well suited to examining policy options that go beyond technology choice,
or hard-to-cost options. E.g. To reduce CO2 you can either (a) use a large
hybrid car, or (b) drive a smaller car.
Relatively complex, opaque and data intensive: hard to apply for less expert
users, so less useful in capacity building efforts.

50

Simulation Models

Simulate behavior of energy consumers and producers under

various signals (e.g. price, income levels, limits on rate of stock
turnover).
Pros:
Not limited by assumption of optimal behavior.
Do not assume energy is the only factor affecting technology choice
(e.g. market share algorithms may be based on both price and
quality of energy service).

Cons:
Tend to be complex and data intensive.
Behavioral relationships can be controversial and hard to
parameterize.
Future forecasts can be sensitive to starting conditions and
parameters.

51

Accounting Frameworks (1)

Physical description of energy system, costs & environmental

impacts optional.

Rather than simulating decisions of energy consumers and

producers, modeler explicitly accounts for outcomes of decisions

So instead of calculating market share based on prices and other

variables, Accounting Frameworks simply examine the
implications of a scenario that achieves a certain market share.

Explores the resource, environment and social cost implications of

alternative future what if energy scenarios.

Example: What will be the costs, emissions reductions and fuel

savings if we invest in more energy efficiency & renewables vs.
investing in new power plants?
52

Accounting Frameworks (2)

Pros:
Simple, transparent & flexible, lower data requirements
Does not assume perfect competition.
Capable of examining issues that go beyond technology
choice or are hard to cost.
Especially useful in capacity building applications.

Cons:
Does not automatically identify least-cost systems: less
suitable where systems are complex and a least cost solution
is needed.
Does not automatically yield price-consistent solutions (e.g.
demand forecast may be inconsistent with projected supply
configuration).
53

Hybrid Models
Many current generation models combine
elements of optimization, simulation and
accounting:
LEAP operates at two levels: basic accounting
relationships are built-in and users can add their
own models on top.
The U.S. National Energy Modeling System (NEMS)
includes optimization modules for the electricity
sector, along with simulation approaches for each
demand sector, all packaged together into a general
equilibrium system.
54

Models vs.
Decision Support Systems
Model methodology is only one (albeit important) issue
for analysts, planners and decision makers.
They also require the full range of assistance provided
by modern decision support systems including: data and
scenario management, reporting, units conversion,
documentation, and online help and support.
Some modern tools such as LEAP focus as much on
these aspects as on the modeling methodology.

55

Tools Compared (1)

Characteristic

LEAP

Developer

Stockholm Environment
Institute

Home page

www.energycommunity.org

Scope

Integrated energy
and GHG scenarios

MARKAL

ENPEP (BALANCE)

MARKAL-MACRO

RETSCREEN

IEA/ETSAP

Natural Resources Canada

www.dis.anl.gov

www.etsap.org

www.retscreen.net

Integrated energy
and GHG scenarios

Integrated energy
and GHG scenarios

Integrated energyeconomy and GHG

scenarios

Accounting & spreadsheet-like mEquilibrium simulation

Accounting
Iteration

Optimization
Linear programming

Hybrid
Accounting
Non-linear programming Accounting

n/a

myopic

Perfect or myopic

Perfect or myopic

Geographic applicability

Local, national, regional, global

Local, national, regional,

global

Local, national, regional, global

Local

Data requirements

Low-medium

Medium-high

Medium-high

Technology specific

Default data included

TED Database with costs,

performace and emission
factors (inc. IPCC factors).
IPCC Emission factors
Coming soon: national energy
& GHG baselines.

None

Extensive defaults: weather

data, products, costs, etc.

Time Horizon

User Controlled. Annual results

User Controlled,
Typically reporting for 5 or 10 year time periods

Primarily static analysis

Methodology
- Model type
- Soution algorithm
- Foresight

Argonne/IAEA

Up to 75 years. Annual
results

Screening of renewable
and CHP projects

n/a

Tools Compared (2)

Characteristic

LEAP

MARKAL/MARKAL-MACRO

ENPEP (BALANCE)

RETSCREEN

Expertise required

Medium

High

High

Low

Level of effort required

Low-Medium

High

High

Low

How Intuitive? (matching

analyst's mental model)

High

Low

Medium

High

Reporting capabilities

Advanced

Basic

Basic

Excel

Data management capabilities

Advanced

Basic

Basic

Excel

Software requirements

Windows

Windows

Windows, GAMS, solver & interface

Excel

Software cost:

Free to NGO, Govt and

researchers in non-OECD
countries.

Free to NGO, Govt

and researchers.

$8,500-$15,000
(including GAMS, solver & interface)

Free

Typical training required

& cost

On request: 5 days/$5000
Also regular international
workshops.

5 days
$10,000

8 days
$30,000-$40,000

Minimal
Free distance learning &
global network of trainers

Technical support
& Cost:

Phone, email or web forum

Free limited support.

Phone or email
$10,000 for 80 hours

Phone or email
$500-$2500 for one year.

Email or web forum

Free limited support.

Reference materials

Manual & training materials

free on web site

Manual available
to registered users

Manual available
to registered users.

Manuals free
on web site

Languages

English, French, Spanish,

Portuguese, Chinese

English

English

Multiple

Long-range Energy
Alternatives Planning System

A software tool for energy planning and climate mitigation scenario

analysis.
Emphasizes ease-of-use, and intuitive and transparent modeling
and data management techniques.
Originally designed for use in developing countries & distributed free
to developing country organizations.
Growing number of users in OECD countries.
Many hundreds of users in over 150 countries.
Widely applied by government energy and environmental agencies,
in academia (for teaching energy and climate policy) in research
institutions and increasingly in energy utilities.
Recently chosen for use by 85 developing countries for use in their
national climate mitigation studies.
www.energycommunity.org

58

Key
Characteristics

An integrated energy-environment, scenario-based modeling system.

Based on simple and transparent accounting and simulation modeling approaches.
Broad scope: demand, transformation, resource extraction, GHG & local air pollutant
emissions, social cost-benefit analysis, non-energy sector sources and sinks.
Used for Forecasting, energy planning, GHG mitigation assessment, emissions
inventories, transport modeling.
Not a model of a particular system, but a tool for modeling different energy systems.
Support for multiple methodologies such as transport stock-turnover modeling,
electric sector load forecasting and capacity expansion and econometric and
simulation models.
Standard energy and emissions accounting built-in. User can also create their own
econometric and simulation models.
Low initial data requirements: most aspects optional.
Includes a Technology and Environmental Database containing costs, performance
and emissions factors of energy technologies, plus IPCC default emission factors.
Links to MS-Office (Excel, Word and PowerPoint).
Local, national, regional and global applicability.
Medium to long-term time frame, annual time-step, unlimited number of years.
Downloadable data sets under development for most countries.
59

LEAP Calculation Flows

MacroEconomics

Demographics

Demand
Analysis

(Pollutant Emissions)

Environmental Loadings

Transformation
Analysis

Stock
Changes

Resource
Analysis

Integrated Cost-Benefit Analysis

Statistical
Differences

Non-Energy Sector
Emissions Analysis

Environmental
Externalities

60

Selected LEAP Studies

APEC Energy
Demand and Supply
Outlook (2006)

Chinas Sustainable
Energy Future
(2003)

Americas Energy
Choices (1991)

Toward a Fossil
Free Energy Future:
The Next Energy
Transition (1992)

Prospectiva
Energetica de
America Latina y el
Caribe (2005)

Implementing
Renewable Energy
Options in South
Africa (2007)

61

More LEAP Applications

USA: Greenhouse gas emissions mitigation in California, Washington, Oregon and

Rhode Island.
Lawrence Berkeley Nat Labs: constructing a global end-use oriented energy model.
Energy and Carbon Scenarios: Chinese Energy Research Institute (ERI) and
LBNL.
Transport Energy Use and Emissions: Various U.S. transportation NGOs (UCS,
ACEEE, SEI) and seven Asian Cities (AIT).
Greenhouse Gas Mitigation Studies: 85 countries are using LEAP for their
UNFCCC National Communications. SEI is assisting the UN to support countries in
this process.
APERC Energy Outlook: Energy forecasts for each APEC economy.
East Asia Energy Futures Project: Study of energy security issues in East Asian
countries including the Koreas, China, Mongolia, Russia, Japan.
Integrated Resource Planning: Brazil, Malaysia, Indonesia, Ghana, South Africa.
Integrated Environmental Strategies: U.S. EPA initiative that engages developing
countries in addressing both local environmental concerns and associated global
greenhouse gas emissions.
City Level Energy Strategies: South Africa.
Sulfur Abatement Scenarios for China: Chinese EPA/UNEP.

62

LEAP Users Map

63

Minimum Hardware & Software

Requirements

Windows 98, 2000, NT, XP, Vista.

400 Mhz Pentium PC, 1024 x 768 screen resolution.
64 MB RAM
Internet Explorer 4.0 or later
Optional: Internet connection, Microsoft Office

64

Status and Dissemination

Available at no charge to non-profit, academic and
governmental institutions based in developing
countries.
Download from http://www.energycommunity.org
Technical support from web site or leap@sei-us.org
User name and password required to fully enable
software. Available on completion of license
agreement.
Most users will need training: available through SEI or
regional partner organizations.
Check LEAP web site for news of training workshops.
65

Typical Data Requirements

M a c ro e c o n o m ic V a ria b le s
S ecto ral d riv in g variab le s
M o re d e ta ile d d riv in g v a ria b le s

G D P /va lue added, population, household size

P roduction of energy in tensive m a te rials (tonnes or $ steel);
transport needs (pass-km , to n n e -km ); in co m e d istrib u tio n , e tc.

E n e rg y D e m a n d D a ta
S ecto r an d su b secto r to ta ls
E n d -u s e a n d te c h n o lo g y
c h a ra c te ris tic s b y s e c to r/s u b s e c to r
P ric e a n d in c o m e re s p o n se (o p tio n a l)

F u e l u se b y se cto r/su b se cto r

a) U sage breakdow n by end-use/device: new vs. existing
buildin gs; vehicle stock by type, vintage; or sim p le r breakdow ns;
b) T e chnology co st and p e rfo rm a n ce
P rice and in co m e elasticities

E n e rg y S u p p ly D a ta
C h a ra c te ris tic s o f e n e rg y s u p p ly ,
tra n s p o rt, a n d c o n v e rsio n fa cilitie s
E n e rg y s u p p ly p la n s
E n erg y reso u rces an d p rices

C apita l and O & M co sts, pe rfo rm a n ce (e fficie n cie s, cap a city

fa cto rs, e tc.)
N e w ca p a city o n -lin e d a te s, co sts, ch a ra cte ristics;
R e se rves of fo ssil fuels; potentia l fo r renew able re so urce s

T e c h n o lo g y O p tio n s
T ec h n o lo g y c o sts a n d p e rfo rm a n c e
P e n e tra tio n ra tes
A d m in is tra tiv e a n d p ro g ram c o s ts
E m is s io n F a c to rs

C apita l and O & M co sts, fo re ig n e xch a n g e , p e rfo rm a n ce

(e fficiency, unit usage, capacity fa ctor, etc.)
P e rcent of new or existing stock replaced per year
E m issions per unit energy consum ed, produced, or transporte d.
66

 An online community with:

discussion & support forums

online libraries and newsletters
downloadable software
training and reference materials

> 2500 members in 150

countries.
www.energycommunity.org
67

Main Screen

68

View
Bar

Analysis View: where you create data structures, enter data, and construct
models and scenarios.

Results View: where you examine the outcomes of scenarios as charts and
tables.

Diagram View: Reference Energy System diagram showing flows of energy

in the area.

Energy Balance: standard table showing energy production/consumption in a

particular year.

Summary View: cost-benefit comparisons of scenarios and other customized

tabular reports.

Overviews: where you group together multiple favorite charts for presentation
purposes.

TED: Technology and Environmental Database technology characteristics,

costs, and environmental impacts of apx. 1000 energy technologies.

Notes: where you document and reference your data and models.

69

70

71

72

The Tree
The main data structure
used for organizing data
and models, and reviewing
results
Icons indicate types of data
(e.g., categories,
technologies, fuels and
effects)
User can edit data
structure.
Supports standard editing
functions (copying, pasting,
drag & drop of groups of
branches)

Tree Branches

Category branches are used mainly for organizing the other branches into hierarchical data
structures.

End-Use branches indicate situations where energy intensities are specified for an
aggregate end-use, rather than with a specific fuel or device. Primarily used when
conducting useful energy analysis.

Technology branches are used to represent final energy consuming devices, and hence
when choosing this type of branch you will also need to select the fuel consumed. The three
basic demand analysis methodologies are represented by three different icons:
Activity Level Analysis, in which energy consumption is calculated as the product of
an activity level and an annual energy intensity (energy use per unit of activity).
Stock Analysis, in which energy consumption is calculated by analyzing the current
and projected future stocks of energy-using devices, and the annual energy intensity of
each device.
Transport Analysis, in which energy consumption is calculated as the product of the
number of vehicles, the annual average distance traveled per vehicle and the fuel
economy of the vehicles.

Key Assumptions branches are used to indicate independent variables (demographic,

macroeconomic, etc.)

In the Transformation tree, fuel branches indicate the feedstock, auxiliary and output fuels
for each Transformation module. In the Resource tree, they indicate primary resources and
secondary fuels produced, imported and exported in your area .
74

Effect branches indicate places where environmental loadings (emissions) are calculated.

Modeling at Two levels

1. Basic physical accounting calculations handled internally within
software (stock turnover, energy demand and supply, electric
dispatch and capacity expansion, resource requirements, costing,
pollutant emissions, etc.).
2. Additional modeling can be added by the user (e.g. user might
specify market penetration as a function of prices, income level
and policy variables).

Users can specify spreadsheet-like expressions that define data and

models, describing how variables change over time in scenarios:
Expressions can range from simple numeric values to complex
mathematical formulae. Each can make use of
1.
2.
3.
4.

math functions,
values of other variables,
functions for specifying how a variable changes over time, or
links to external spreadsheets.

75

Top-Level Tree Categories

Key Assumptions: independent variables (demographic, macroeconomic,

etc.)
Demand: energy demand analysis (including transport analyses).
Statistical Differences: the differences between final consumption values
and energy demands.
Transformation: analysis of energy conversion, extraction, transmission
and distribution. Organized into different modules, processes and output
fuels.
Stock Changes: the supply of primary energy from stocks. Negative values
indicate an increase in stocks.
Resources: the availability of primary resources (indigenous and imports)
including fossil reserves and renewable resources.
Non-energy sector effects: inventories and scenarios for non-energy
related effects.

76

Expressions
Similar to expressions in spreadsheets.
Used to specify the value of variables. Expressions can be
numerical values, or formulae that yield different results in each
scenario year.
Can use many built-in functions, or refer to the values of other
variables.
Can be linked to Excel spreadsheets.
Inherited from one scenario to another.

77

Some Expression Examples

Simple Number

Simple Formula

Example: 0.1 * 5970

Growth Rate

Calculates a constant value in all scenario years.

Example: Growth(3.2%)
Calculates exponential growth over time.

Interpolation Function

Example: Interp(2000, 40, 2010, 65, 2020, 80)

Calculates gradual change between data values

Step Function

Example: Step(2000, 300, 2005, 500, 2020, 700)

Calculates discrete changes in particular years
GrowthAs

Example: GrowthAs(Income,elasticity)
Calculates future years using the base year value of the current
branch and the rate of growth in another branch.

Many others!
78

Four Ways to Edit an Expression:

Type to directly edit the

expression.
Select a common function from a
selection box.
Use the Time-Series Wizard to
enter time-series functions (Interp,
Step, etc. and to link to Excel)
Use the Expression builder to
make an expression by draggingand-dropping functions and
variables.

79

Scenarios in LEAP

Consistent story-lines of how an energy system might evolve over

time. Can be used for policy assumption and sensitivity analysis.
Inheritance allows you to create hierarchies of scenarios that inherit
default expressions from their parent scenario. All scenarios inherit
from Current Accounts minimizing data entry and allowing common
assumptions to be edited in one place.
Multiple inheritance allows scenarios to inherit expressions from
more than one parent scenario. Allows combining of measures to
create integrated scenarios.
The Scenario Manager is used to organize scenarios and specify
inheritance.
Expressions are color coded to show which expressions have been
entered explicitly in a scenario (blue), and which are inherited from a
parent scenario (black) or from another region (purple).

80

The Scenario Manager

81

Demand Analysis in LEAP

Analysis of energy consumption and associated
costs and emissions in an area.
Demands organized into a flexible hierarchical
tree structure.
Typically organized by sector, subsector, enduse and device.
Supports multiple methodologies:
End-use analysis: energy = activity level x energy
intensity
Econometric forecasts
Stock-turnover modeling

82

Demand Modeling Methodologies

1. Final Energy Analysis: e = a i

Where e=energy demand, a=activity level, i=final energy

intensity (energy consumed per unit of activity)
Example: energy demand in the cement industry can be
projected based on tons of cement produced and energy used
per ton. Each can change in the future.

2. Useful Energy Analysis: e = a (u / n)

Where u=useful energy intensity, n = efficiency

Example: energy demand in buildings will change in future as
more buildings are constructed [+a]; incomes increase and so
people heat and cool buildings more [+u]; or building insulation
improves [-u]; or as people switch from less efficient oil boilers
to electricity or natural gas [+n].
83

Demand Modeling Methodologies

(2)
3. Transport Stock Turnover Analysis: e = s m /
fe
Where: s= number of vehicles (stock),
m = vehicle distance, fe = fuel economy
Allows modeling of vehicle stock turnover.
Also allows pollutant emissions to be
modeled as function of vehicle distance.
Example: model impact of new vehicle fuel
economy or emissions standards.
84

A Simple Demand Data Structure

Households
(8 million)

Urban
(30%)

Electrified
(100%)

Lighting
(100%)

Existing (80%, 400 kWh/yr)

Efficient (20%, 300kWh/yr)

Refrigeration
(80%)

Rural
(70%)

Electrified
(20%)

Cooking
(100%)
Other
(50%)

Non-Electrified
(80%)

The tree is the main data structure used for organizing data
and models, and for reviewing results.
Icons indicate the types of data (e.g., categories,
technologies, fuels and environmental effects).
Users can edit the tree on-screen using standard editing
functions (copy, paste, drag & drop)
Structure can be detailed and end-use oriented, or highly
aggregate (e.g. sector by fuel).
Detail can be varied from sector to sector.

Transformation Analysis in
LEAP

Analysis of energy conversion, transmission and distribution, and

resource extraction.
Demand-driven engineering-based simulation.
Basic hierarchy: modules (sectors), each containing one or more
processes. Each process can have one or more feedstock fuels
and one or more auxiliary fuels.
Allows for simulation of both capacity expansion and process
dispatch.
Calculates imports, exports and primary resource requirements.
Tracks costs and environmental loadings.

86

Standard Transformation
Module

Auxiliary Fuel Use

Output
Fuel

Process
(efficiency)

Output
Fuel

Process
(efficiency)
Module
Dispatch

Output
Fuel

Process
(efficiency)

Output
Fuel

Process
(efficiency)

Output
Fuel

Process
(efficiency)

Feedstock Fuel
Feedstock Fuel

Feedstock Fuel
Feedstock Fuel

Feedstock Fuel
Feedstock Fuel

Feedstock Fuel
Feedstock Fuel

Feedstock Fuel
Feedstock Fuel

Auxiliary Fuel Use

Co-Product
Fuel (e.g Heat)

87

Simple Transformation Module

88

Transformation Modules with Feedbacks

89

Electric Generation Simulation

Examines both Capacity Expansion (MW) and Plant
Dispatch (MWh).
Exogenous Capacity: User specifies current and possible future
capacity of plants (MW)
Endogenous Capacity: User specifies types of plants to be built
but LEAP decides when to add plants to maintain a specified
planning reserve margin.

Two Modes of Dispatch simulation:

Mode 1: Historical: LEAP simply dispatches plants based on
historical generation.
Mode 2: Simulation: plants dispatched based on various
dispatcxh rules ranging from very simple (% of total generation)
to quite sophisticated (dispatch in order of running costs)
90

Electric Generation (2)

Plants are dispatch to meet both total demand (in MWh)
as well as the instantaneous peak demand which varies
from by hour, day and season.
User can exogenously specify a load-duration curve and
LEAP will dispatch plants by merit order.
Alternatively, load shapes be specified for each demand
device so that the overall system load is calculated
endogenously. Thus the effect of DSM policies on the
overall load shape can then be explored in scenarios.
Plant dispatch can also then be varied by season (e.g. to
reflect how hydro dispatch may vary between wet and
dry seasons).
91

Hourly Demand Curve

Hour-by-hour load curve
Power demand in each hour of the year
Area = Power (kW) x time (1 hour) = Energy (kWh)

hour number in year

10

8759 8760

Load Duration Curve

Rearrange hourly demand curve
Hours on x-axis is # of hours/year that demand is greater
than or equal to a particular value

hour number in year

10

8759 8760

Load-Duration Curve and

System Dispatch in LEAP
100
95

Peak Load
Plants

90
85

Percent of Peak Load

80
75
70
65

Intermediate
Load Plants

60
55
50
45
40
35
30

Baseload
Plants

25
20

Capacity (MW) * MCF

15
10
5
0
0

500

1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 6,500 7,000 7,500 8,000 8,500
Hours Sorted from Highest to
Lowest Demand

94

Oil Refining Simulation

Uses the same basic module structure as for Electric
Generation, but generally have a single input fuel (crude)
and multiple output fuels (gasoline, diesel, kerosene,
LPG, fuel oil , etc.)
Outputs produced in specified proportions, and the
whole module is run to the point where demands for
priority products are met (assuming module has
sufficient capacity).
Other products are considered by-products and may or
may not be produced in sufficient quantities.
User sets simulation rules to tell what LEAP to do in
situations of surpluses (export or waste) and deficits
(import or ignore).
95

Simple Refinery Simulation Example

96

Emissions Accounting

Emission factors for any greenhouse gas or local air pollutant can be
entered in LEAP and used to calculated emissions loadings for any
scenario.
Factors can be specified in any physical unit and can be
denominated by units of either energy consumption or production
(e.g. kg/ton of coal) or distance driven for transport factors (e.g.
grams/mile).
Emission factors can also be specified in terms of the chemical
composition of fuels (e.g. sulfur) so that factors can be corrected if
fuel composition is different from the default in the area of study
(e.g. if a country has high sulfur coal).
LEAP can use emission factors entered in the accompanying TED
database which includes all of the default IPCC GHG emission
factors.
Emission results can be shown for individual pollutants or summed
across all greenhouse gases in terms of the overall Global Warming97
Potentials (GWPs).

TED:
The Technology and Environmental
Database
Fields
Information
Pages

Technology
Data

Cost
Data

Environmental Notes
Reference
Impacts

Technologies
Demand

Conversion

Database Contents

Supply:
Extraction
Transmission &
Distribution
98

Social CostBenefit Analysis in

LEAP
Societal perspective of costs
and benefits (i.e. economic not
financial analysis).
Avoids double-counting by
drawing consistent boundary
around analysis (e.g. whole
system including.
Cost-benefit analysis calculates
the Net Present Value (NPV) of
the differences in costs between
two scenarios.
NPV sums all costs in all years
of the study discounted to a
common base year.
Optionally includes externality
costs.

Demand
(costs of saved energy,
device costs, other non-fuel
costs)

Transformation
(Capital and O&M costs)

Primary Resource Costs

or
Delivered Fuel Costs

Environmental
Externality Costs
99

Simple Cost-Benefit Analysis

Example
Two scenarios for meeting future growth in electricity lighting demand:

1. Base Case

Demand: future demand met by cheap incandescent bulbs.

Transformation: growth in demand met by new fossil fired
generating capacity.

2. Alternative Case

Demand: DSM programs increase the penetration of efficient

(but more expensive) fluorescent lighting.
Transformation: Slower growth in electricity consumption and
investments to reduce transmission & distribution losses mean
that less generating capacity is required.
100

Simple Cost-Benefit Analysis (cont.)

The Alternative Case

uses more expensive (but longer lived) lightbulbs.
Result: depends on costs, lifetimes, & discount rate.

requires extra capital and O&M investment in the electricity

transmission & distribution system.
Result: net cost

..requires less generating plants to be constructed (less capital and

O&M costs).
Result: net benefit

requires less fossil fuel resources to be produced or imported.

Result: net benefit

produces less emissions (less fuel combustion).

Result: net benefit (may not be valued)
101

Energy Balances
An accounting system that describes the flows of energy through an
economy, during a given period.

Imports

Non-energy consumption
Transformation Sectors
(e.g. petrochemical
Losses and Consumption feedstock, fertilizers)

P + I X = L + CF + CNE + DS
Total Primary
Energy Produced

Exports

Total Final Energy

Use in Consuming
Sectors

Net Changes
in Stocks

Sample IEA Energy Balance

Breakdown by
Sector and
Activities
Breakdown
by Energy
Source

Energy Balances in LEAP

Results automatically formatted as standard energy
balance tables in Energy Balance View.
Balances can be viewed for any year, scenario and
region in different units.
Balance columns can be switched between fuels, fuel
groupings, years, and regions.
Balance rows are Demand sectors and Transformation
modules.
Display in any energy unit.
Balance can also be shown in chart or energy flow
diagram formats.
104

LEAP Energy Balance Table

105

LEAP Energy Balance Diagram

106

Multi-Region Analysis
LEAP supports multi-region analyses.
Regions appear as an extra data dimension.
Each region shares a similar basic tree structure
although tree branches can be selectively
hidden in different regions.
All results can be summed and displayed across
regions or aggregated into groups of regions
Forthcoming: LEAP 2007 will support interregional trade calculations so that import
requirements for some regions will drive
107
production and exports in other regions.

Showing Results for a MultiRegion Data Set in LEAP

108

The Application Programming

Interface (API)
LEAPs API is a standard COM Automation Server
Other programs can control LEAP: changing data
values, calculating results, and exporting them to Excel
or other applications.
For example, a script could iteratively run LEAP multiple
times revising input assumptions for goal-seeking
applications.
LEAP has a built-in script editor that can be used to edit,
interactively debug and run scripts that use its API.
LEAP uses Microsoft's ActiveScript technology which
supports in Visual Basic and JavaScript.
109

LEAP Terminology

Area: the system being studied (e.g. country or region).

Current Accounts: the data describing the Base Year (first year) of the study period.
Scenario: one consistent set of assumptions about the future, starting from the
Current Accounts. LEAP can have any number of scenarios. Typically a study
consists of one baseline scenarios (e.g. business as usual) plus various counterfactual policy scenarios.
Tree: the main organizational data structure in LEAP a visual tree similar to the one
used in Windows Explorer.
Branch: an item on the tree: branches can be organizing categories, technologies,
modules, processes, fuels and independent driver variables, etc.
Views: The LEAP software is structured as a series of different views onto an
energy system.
Variable: data at a branch. Each branch may have multiple variables. Types of
variables depend on the type of branch, and its properties. In LEAP, Variables are
displayed as tabs in the Analysis view.
Disaggregation: the process of analyzing energy consumption by breaking down
total demand into the various sectors, subsectors, end-uses and devices that
consume energy.
Expression: a mathematical formula that specifies the values of a variable over time
at a given branch and for a given scenario. Expressions can be simple values, or 110
mathematical formula that yield different results in different years

When you have a problem

Post message on LEAP forum at
www.energycommunity.org or email leap@sei-us.org
BE AS SPECIFIC AS POSSIBLE
Include:
Error message (if any)
Did problem happen during installation or when running LEAP?
What were you doing and what part of LEAP were you using
when problem occurred?
Is the problem reproducible and what steps do I need to take do
that?
Operating system version (2000, XP, Vista, etc.) and language
Version of LEAP (check Help: About)
If possible include the LEAP.LOG file and attach the problem
data set as a zip file.
111

Part 3: LEAP Demonstration

Exercises

Two sessions on Wednesday (26th) at 13-17 (group A) and Thursday

(27th) at 8:00-12.00 (group B) in Trotter (Trtter, Brinellvg 64).
You will work in groups of 2-3 per PC.
Each session is a 2 part exercise:
Part 1: A screening exercise: using Excel to do simple screening of
various GHG mitigation options.
You will be provided with a partly completed Excel spreadsheet that you will
need to complete.
Work in groups to develop a screening matrix.

Part 2: Preparing an integrated mitigation assessment using LEAP (and

options assessed in part 1)

Lab evaluation: this will be described in the lab.

113

An online initiative designed to

foster a community among analysts
working on energy and
sustainability issues.
Managed by SEI in collaboration
with regional partners in Africa,
Europe and Latin America.
Open to all at no charge.
Activities:
Annual regional training workshops
in Africa & Latin America.
The COMMEND web site
Technical support for energy
analysts in developing countries.
Development, maintenance and
tech support for SEIs LEAP
software.
Semi-annual newsletter
(reCOMMEND)
114