getting started with times-veda version 2 - … · getting started with times-veda version 2.7 may...

ENERGY TECHNOLOGIES SYSTEMS ANALYSIS PROGRAMME

GETTING STARTED WITH TIMES-VEDA

Version 2.7 May 2009

By: MAURIZIO GARGIULO

Version 2.7 – May 2009

2

Acknowledgment

The author is grateful for the useful suggestions and comments by Gary Goldstein,

Antti Lehtila and Uwe Remme to the first version (February 2008) and by Amit

Kanudia and GianCarlo Tosato to this second version.


3

Table of Contents

1 Introduction .................................................................................................................................. 9 2 How to model a simplified energy system: step-by-step description ........................................ 11

2.1 A simplified energy system and its possible developments ............................................... 11 2.1.1 Description of the example system and analysis questions ........................................... 11

2.1.2 ............................................................................................................................................ 12 2.1.3 Description in TIMES: base elements ........................................................................... 12

2.1.3.1 Commodities and processes ................................................................................... 12 2.1.3.2 Possible system development: Scenarios ............................................................... 13 2.1.3.3 Representation of the Reference Energy System ................................................... 13

2.1.3.4 The time dimension ................................................................................................ 15 2.1.3.5 Calibration and Establishing the Reference Scenario ............................................ 15

2.1.3.6 Scenarios ................................................................................................................ 16 2.2 Declaration of the model through VEDA-FE .................................................................... 16

2.2.1 Overview of VEDA-FE workbooks and internal syntax ............................................... 16 2.2.1.1 VEDA template controls ........................................................................................ 16

2.2.1.2 VEDA rules ............................................................................................................ 17 2.2.1.3 VEDA workbooks .................................................................................................. 17

2.2.2 Sub-folders structure ...................................................................................................... 18 2.2.3 SysSettings: Model Setup Template .............................................................................. 20

2.2.3.1 Declaring regions and time-slices .......................................................................... 21

2.2.3.2 Declaring model horizon ........................................................................................ 21

2.2.3.3 Declaring settings for the synchronization process................................................ 22 2.2.3.4 Declaring inter/extrapolation options and dummy import prices .......................... 23 2.2.3.5 Declaring Constants ............................................................................................... 24

2.2.4 Commodities and processes definition in the base year template file

VT_TT_SUP_V1p0 ................................................................................................................... 24

2.2.4.1 Balance Sheet ......................................................................................................... 25 2.2.4.2 SUP_Comm Sheet.................................................................................................. 26 2.2.4.3 SUP_Process Sheet ................................................................................................ 27

2.2.5 Process characterization file VT_TT_SUP_V1p0 ......................................................... 28 2.2.5.1 MIN Sheet .............................................................................................................. 29

2.2.5.2 IMP-EXP Sheet ...................................................................................................... 30 2.2.5.3 DEMAND Sheet .................................................................................................... 31 2.2.5.4 EMIssion Sheet ...................................................................................................... 32

2.2.6 SuppXlS template folder ................................................................................................ 33 2.2.6.1 Scenario files .......................................................................................................... 34

2.2.6.2 Demand sub-folder ................................................................................................. 34 2.2.6.3 Trades sub-folder ................................................................................................... 34 2.2.6.4 UConstraints sub-folder ......................................................................................... 35

2.2.7 SubRES_TMPL templates folder................................................................................... 35 2.2.7.1 SubRES-B-NewTechs ............................................................................................ 35

2.2.7.2 SubRES-B-NewTechs_Trans................................................................................. 35 2.3 Managing the Templates via VEDA-FE ............................................................................ 36

2.3.1 The VEDA-FE Navigator .............................................................................................. 36 2.3.2 Select the model to be processed ................................................................................... 36 2.3.3 Synchronize model templates and internal database ...................................................... 37 2.3.4 Browse/Edit data in VEDA-FE database ....................................................................... 39

2.3.4.1 Search and view data .............................................................................................. 39


4

2.3.4.2 Process and Commodity Master ............................................................................ 45

2.3.4.3 Commodity Group Master ..................................................................................... 46 2.3.4.4 RES view................................................................................................................ 46

2.4 Generating and solving the TIMES model via VEDA-FE ................................................ 47 2.4.1 Running and solving the model ..................................................................................... 47 2.4.2 Generation and solution files ......................................................................................... 50

2.4.2.1 Contents of the run folders ..................................................................................... 50 2.4.2.2 Data Dictionary files .............................................................................................. 51 2.4.2.3 LST file .................................................................................................................. 54 2.4.2.4 Log files ................................................................................................................. 59 2.4.2.5 VEDA-BE solution files ........................................................................................ 60

2.5 Analyzing the model results with VEDA-BE .................................................................... 60 2.5.1 Results import in VEDA-BE.......................................................................................... 60

2.5.2 Results tables.................................................................................................................. 62

2.5.2.1 Check dummy imports ........................................................................................... 62 2.5.2.2 All prices ................................................................................................................ 63 2.5.2.3 Demands................................................................................................................. 64 2.5.2.4 Total supply............................................................................................................ 64 2.5.2.5 Creating a new table - Emissions ........................................................................... 65

2.5.2.6 All costs.................................................................................................................. 67 2.5.2.7 Total System Cost .................................................................................................. 68

2.5.3 Evaluating the effect of possible choices ....................................................................... 69

2.5.4 Setting Defaults .............................................................................................................. 70 2.5.5 Exporting result tables ................................................................................................... 70

2.5.6 Graphs ............................................................................................................................ 71 3 Next Step: the TIMES_Demo Model ......................................................................................... 72

3.1 Main aspects of the TIMES_DEMO model ....................................................................... 72 3.1.1 Installation of the model ................................................................................................ 72 3.1.2 Directory structure ......................................................................................................... 74

3.1.3 Description of the system represented by TIMES_DEMO ........................................... 75 3.2 How to represent in the model the main components of an energy system ....................... 76

3.2.1 How to declare regions, time horizon and time slices ................................................... 76 3.2.1.1 Time slices description........................................................................................... 77 3.2.1.2 Declaring Commodity Group ................................................................................. 78

3.2.2 How to declare commodities.......................................................................................... 78

3.2.2.1 General features ..................................................................................................... 78 3.2.2.2 Emissions ............................................................................................................... 79

3.2.2.3 Materials................................................................................................................. 79 3.2.3 How to declare a process: general features .................................................................... 80

3.2.3.1 Definition of process activity variables ................................................................. 80 3.2.3.2 Use of capacity ....................................................................................................... 81 3.2.3.3 Defining flow relationships in a process ................................................................ 81

3.2.3.4 Limiting flow shares in flexible processes ............................................................. 82 3.2.3.5 Peaking Reserve Requirements (time-sliced commodities only) .......................... 82

3.2.4 How to declare specific processes ................................................................................. 83 3.2.4.1 Mining process and import/export processes ......................................................... 83 3.2.4.2 Flexible refinery ..................................................................................................... 83

3.2.4.3 Electric power plants .............................................................................................. 84 3.2.4.4 Cogeneration power plant ...................................................................................... 86

3.2.4.5 District heating plants ............................................................................................ 90 3.2.4.6 Cars, trains, bus – converting activity, capacity, demand units; and dual-purpose

devices 90


5

3.2.4.7 Car process- defining demand and the load shape ................................................ 91

3.2.4.8 Industrial process ................................................................................................... 92 3.2.5 How to construct a scenario file ..................................................................................... 93 3.2.6 How to construct User Constraints (UC) ....................................................................... 94

3.2.6.1 User Constraints (UC) in Scenario Files ................................................................ 95 3.2.6.2 How to construct special flow share scenarios ...................................................... 95

3.2.6.3 The UC scenario file contains the following sheets: .............................................. 96 3.2.7 How to construct a new technologies file (SubRes) ...................................................... 97

3.2.7.1 How to construct a SubRes file .............................................................................. 97 3.2.7.2 How to construct a SubRes transformation file ..................................................... 98

3.2.8 How to construct a demand file ..................................................................................... 98

3.2.9 How to construct a trade scenario .................................................................................. 99 3.2.9.1 How to declare a trade matrix ................................................................................ 99

3.2.9.2 How to declare a trade parameters ....................................................................... 100

3.3 Interpolation and extrapolation ........................................................................................ 100 3.3.1 Defaults inter/extrapolation .......................................................................................... 101 3.3.2 Enhanced user-controlled interpolation / extrapolation ............................................... 102 3.3.3 Interpolation of cost parameters ................................................................................... 103

4 Appendix A - Getting Started with Problem – Defining and Describing the Area of Study ... 105

4.1 The multiple dimensions of energy systems .................................................................... 105 4.1.1 Energy: from primary resource to end-use services .................................................... 106 4.1.2 Engineering: technology efficiency and system efficiency ......................................... 108

4.1.3 Economics: the value of energy systems ..................................................................... 111 4.1.4 Emissions and the environment ................................................................................... 112

4.2 The systems analysis approach: identification of the areas of study ............................... 113 4.2.1 Scope of the analysis .................................................................................................... 113

4.2.2 Boundaries ................................................................................................................... 115 4.2.3 The time dimension ...................................................................................................... 115

4.2.3.1 Time horizon ........................................................................................................ 115

4.2.3.2 Time granularity ................................................................................................... 116 4.2.4 Components ................................................................................................................. 119

4.2.4.1 Commodities ........................................................................................................ 119 4.2.4.2 Technologies ........................................................................................................ 119

4.2.5 Connections: the Reference Energy System ................................................................ 120 4.3 The systems analysis approach: quantification ................................................................ 123

4.3.1 Flows of energy commodities ...................................................................................... 123 4.3.2 Energy technology and end-use devices ...................................................................... 126

4.3.3 Emissions ..................................................................................................................... 128 4.3.4 Quantifying the economic dimension of the system .................................................... 130

4.4 The systems analysis approach: control ........................................................................... 134 4.4.1 Preparation of the mental experiments......................................................................... 135

4.4.1.1 Representation of the system in a model ............................................................. 135

4.4.1.2 Uncontrollable and controllable exogenous variables ......................................... 137 4.4.1.3 Objectives and targets .......................................................................................... 139 4.4.1.4 Policies instruments and specific measures as control variables ......................... 139

4.4.2 Carrying out the mental experiments: scenarios .......................................................... 141 4.4.3 Robust and hedging strategies ...................................................................................... 143

5 References ................................................................................................................................ 145


6

List of tables Table 2.1. Quantitative input data of the simplified energy supply sub-system ................................ 12 Table 2.2 Data Dictionary files for the TIMES Tutorial example (to be read by column) ................ 51 Table 2.3 Equation Listing ................................................................................................................. 55 Table 2.4 Solution Report .................................................................................................................. 58

Table 2.5 Solution list ........................................................................................................................ 58 Table 2.6 Some scenario indicators ................................................................................................... 69 Table 3.1: Headers in the scenario files to define subsets of technologies/commodities .................. 94 Table 2.7: Parameters not being inter/extrapolated by default ........................................................ 101 Table 2.8: Option codes for the control of data interpolation .......................................................... 103

Table 4.1: Global summary energy balance for 1973 [IEA, 2004] (a) ............................................ 124

Table 4.2: Global summary energy balance for 2002 [IEA, 2004] (a) ............................................ 125

Table 4.3: Example of data needed for characterising energy technologies .................................... 127

List of figures

Figure 1-1 ETSAP tools and typical applications .............................................................................. 10 Figure 2-1 The Reference Energy System of the above simplified energy ....................................... 14

Figure 2-2 VEDA_Models Folder ..................................................................................................... 19 Figure 2-3 Folder TIMES Tutorial..................................................................................................... 19 Figure 2-4 Sub-folder SubRES_TMPL.............................................................................................. 20

Figure 2-5 Sub-folder SuppXLS ........................................................................................................ 20

Figure 2-6 SysSettings: Example 1 – Regions and Time slices related declarations ........................ 21 Figure 2-7 SysSettings: Example 2 – Model horizon declarations .................................................... 21 Figure 2-8 SysSettings: Import Settings ............................................................................................ 22

Figure 2-9 SysSettings – Inter_Extrapolation and Default options sheet .......................................... 23 Figure 2-10 SysSettings - Constants sheet ......................................................................................... 24

Figure 2-11 Balance sheet of VT_TT_SUP_V1p0 ............................................................................ 25 Figure 2-12 SUP_Comm sheet of VT_TT_SUP_V1p0 ..................................................................... 26

Figure 2-13 SUP_Process sheet of VT_TT_SUP_V1p0 ................................................................... 27 Figure 2-14 Topology and parameter defnition table ........................................................................ 28 Figure 2-15 MIN sheet of VT_TT_SUP_V1p0 ................................................................................. 29

Figure 2-16 IMP-EXP sheet of VT_TT_SUP_V1p0 – IMPOILCRD1 ............................................. 30 Figure 2-17 IMP-EXP sheet of VT_TT_SUP_V1p0 – Emission Permits CO2N ............................. 31 Figure 2-18 Demand sheet of VT_TT_SUP_V1p0 ........................................................................... 32 Figure 2-19 Emission sheet of VT_TT_SUP_V1p0 .......................................................................... 33

Figure 2-20 SuppXlS folder ............................................................................................................... 33 Figure 2-21 Scen_CO2N_Bound ....................................................................................................... 34 Figure 2-22 Scen_IMP_Bound .......................................................................................................... 34 Figure 2-23 The VEDA-FE Navigator............................................................................................... 37 Figure 2-24 Active models in VEDA-Navigator ............................................................................... 37

Figure 2-25 The VEDA Front End Navigator ................................................................................... 38 Figure 2-26 Processing templates ...................................................................................................... 38

Figure 2-27 Templates and databases are consistent ........................................................................ 38 Figure 2-28 VEDA-FE Browse/Edit .................................................................................................. 39 Figure 2-29 VEDA-FE TIMES View Browse ................................................................................... 40 Figure 2-30 Example of dynamic data cube (pivot table) .................................................................. 40 Figure 2-31 Example of dynamic data cube rearrange ...................................................................... 41


7

Figure 2-32 VEDA-FE Search Engine - Processes ............................................................................ 42

Figure 2-33 Example of filter use in VEDA-FE browse ................................................................... 42 Figure 2-34 Example of direct edit data from VEDA-FE browse ..................................................... 43 Figure 2-35 How to change a data from direct edit ........................................................................... 43 Figure 2-36 FE Navigator after the direct edit ................................................................................... 43 Figure 2-37 FE Navigator after the edit ............................................................................................. 44

Figure 2-38 The base year template updated ..................................................................................... 44 Figure 2-39 Process Master ................................................................................................................ 45 Figure 2-40 Commodity Master ......................................................................................................... 45 Figure 2-41 Commodity Group Master ............................................................................................. 46 Figure 2-42 RES view in VEDA-FE.................................................................................................. 46

Figure 2-43 RES view OILCRD ........................................................................................................ 47 Figure 2-44 VEDA-FE Case Manager for submitting model runs .................................................... 47

Figure 2-45 Cplex options.................................................................................................................. 48

Figure 2-46: Default value for LIMROW and LIMCOL ................................................................... 49 Figure 2-47: Command window showing the beginning and the optimal solution ........................... 49 Figure 2-48: Set Model Variants ........................................................................................................ 50 Figure 2-49 Files in the VEDA-FE Log folder .................................................................................. 59 Figure 2-50 VEDA-BE results import ............................................................................................... 61

Figure 2-51 Path database TIMES_TUTORIAL ............................................................................... 61 Figure 2-52 VEDA-BE Import/Archive window............................................................................... 61 Figure 2-53 VEDA-BE main table specification form ...................................................................... 62

Figure 2-54 Prices by commodity, scenario and year ........................................................................ 64 Figure 2-55 Demands by process, year and scenario ......................................................................... 64

Figure 2-56 Total supply .................................................................................................................... 65 Figure 2-57 Selecting the VAR_FOUT attribute ............................................................................... 66

Figure 2-58 Selecting Commodity Set ENV ...................................................................................... 66 Figure 2-59 Emissions by source, year and scenario ......................................................................... 66 Figure 2-60 All costs table by variable, scenario and time period ..................................................... 67

Figure 2-61 _SysCost Table ............................................................................................................... 68 Figure 2-62 VEDA-BE setting options .............................................................................................. 70

Figure 3-1 VEDA-FE Navigator and Case Manager forms of the TIMES_DEMO model ............... 73 Figure 3-2 Command window ........................................................................................................... 73 Figure 3-3 TIMES_DEMO directory ................................................................................................. 75 Figure 3-4 SysSettings: Demo model – Regions and Timeslices related declarations ...................... 76

Figure 3-5 SysSettings: Demo model - Time periods related declarations ........................................ 76 Figure 3-6 Times slices levels ............................................................................................................ 77

Figure 3-7 SysSettings: Demo model – Commodities Group ............................................................ 78 Figure 3-8 Example of coefficients for combustion emissions.......................................................... 79 Figure 3-9 Flexible Refinery .............................................................................................................. 84 Figure 3-10 Electricity power plants .................................................................................................. 86 Figure 3-11 Back pressure turbine characteristic curve ..................................................................... 87

Figure 3-12 Condensing combined heat and power characteristic curve .......................................... 88 Figure 3-13 Condensing combined heat and line fuel ....................................................................... 89 Figure 3-14 Cogeneration power plants ............................................................................................. 89 Figure 3-15 District heating plants..................................................................................................... 90 Figure 3-16 Cars................................................................................................................................. 91

Figure 3-17 Cars demand and load shape .......................................................................................... 91 Figure 3-18 Industrial processes – Iron and steel pellet and sinter production .................................. 92

Figure 3-19 Industrial processes – Raw iron production ................................................................... 92 Figure 3-20 Industrial processes – Crude steel production ................................................................ 92 Figure 3-21 Industrial processes – Iron and Steel production ........................................................... 93


8

Figure 3-22 New Scenario file from VEDA-FE Navigator ............................................................... 93

Figure 3-23 Scenario file from the Demo .......................................................................................... 94 Figure 3-24 User Constraint Scenario in VEDA Navigator .............................................................. 97 Figure 3-25 Example of new electricity power plant in the SubRes ................................................. 97 Figure 3-26 Example of new industry technologies in the SubRes ................................................... 98 Figure 3-27 Demand driver ................................................................................................................ 98

Figure 3-28 Demand sensitiviy series ................................................................................................ 99 Figure 3-29 Demand drivers allocation table ..................................................................................... 99 Figure 3-30 Trade matrix declaration .............................................................................................. 100 Figure 3-31 Trade parameters .......................................................................................................... 100 Figure 4-1: The energy system: schematic diagram with some illustrative examples of the energy

sector and energy end-use and services. .......................................................................................... 107 Figure 4-2: Major energy and carbon flows through the global energy systems in 1990................ 109

Figure 4-3: - Example of load curve profile for district heating production .................................... 117

Figure 4-4: Monthly average load curve for district heating production ......................................... 117 Figure 4-5: Duration curve for district heating production .............................................................. 118 Figure 4-6: Example of RES – Reference Energy System .............................................................. 121 Figure 4-7 Partial view of a simple Reference Energy System ....................................................... 123 Figure 4-8 Split of fuel consumption ............................................................................................... 126

Figure 4-9: Simplified characterisation of existing Power Plants in Tuscany, Italy........................ 128 Figure 4-10: Simplified characterisation of new Power Plant in Italy ............................................. 128 Figure 4-11 Source: CORINAIR – Emissions by category ............................................................. 129

Figure 4-12 Source: U.S. EPA 1998 – Emissions by category ........................................................ 130 Figure 4-13: Fuel Retail Prices, 1/2 (US$/Unit) .............................................................................. 131

Figure 4-14: Fuel Retail Prices, 2/2 (US$/Unit) .............................................................................. 132 Figure 4-15 Example of step-wise (inverse) supply-demand curves for electricity ........................ 133

Figure 4-16: Selection of scenarios in an events tree: an example .................................................. 142 Figure 4-17: The event tree of possible climate sensitivity values .................................................. 143 Figure 4-18: Example of exploratory scenarios without a hedging strategy ................................... 144

Figure 4-19: Example of hedging strategy ....................................................................................... 144


9

1 Introduction

This manual explains how to start building the technical-economic model of your energy system,

and its possible developments over time, with TIMES-VEDA. TIMES (The Integrated MARKAL /

EFOM System1) is the most recent model generator

2 of ETSAP

3. VEDA (VErsatile Data Analyst) is

the data management and analysis system (“shell”)4.

This Users‟ Guide is addressed to beginners who have some theoretical knowledge about energy

systems in general. Otherwise it can be helpful to read Appendix A, a primer on energy systems

analysis. This guide illustrates step-by-step how to build an energy model, from the simplest case

with one commodity and one technology to a complex model encompassing the entire energy

system with dozens of commodities and hundreds of technologies.

It looks to provide the foundation for the development of an appropriate TIMES model able to

support in-depth qualitative studies and quantitative analyses of the energy system whose possible

future developments you want to study.

To effectively use this tutorial you should have TIMES and VEDA installed on your PC and

running correctly. You can verify this by running the Demo model, which comes with the

installation, and check that the objective function matches the value reported for the run in the

installation documentation at www.kanors.com/vedasupport. If you don‟t have the software yet,

please follow the instruction given by http://www.etsap.org/Tools.asp to get it. If the software is not

yet installed, or the Demo problem does not run correctly, refer to www.kanors.com/vedasupport

(installation).

The same web site gives further details for using VEDA.

Chapter 2 illustrates the main steps necessary to represent a simplified energy system:

Identification of the system and description of its possible development paths;

Collection of statistical data and definition of energy network and its technologies in a series

of Excel workbooks;

Import of the templates into VEDA5 Front End 4 for browsing the model data;

1 To know more about the TIMES see the user guidebook at http://www.etsap.org/documentation.asp.

2 The model generator is a program written in the General Algebraic Modeling System (GAMS) specialized

programming language. It reads user data and produces a matrix of coefficients that defines a linear programming

problem, which is then seamlessly passed to an appropriate solver. Model results are then assembled by a report writer

also written in GAMS. 3 The Energy Technology Systems Analysis Program of the International Energy Agency (IEA/ETSAP) is a

longstanding international research agreement among organizations of more than 20 OECD countries. 4 There is also a version of the ANSWER user support system that has evolve from MARKAL to support TIMES. A

supplement for ANSWER-TIMES is being prepared. 5 To know more about VEDA see at http://www.kanors.com/vedasupport

http://www.kanors.com/vedasupport

http://www.etsap.org/Tools.asp


http://www.etsap.org/documentation.asp



10

Generation of TIMES optimization problem in GAMS format, and running the model;

Analyzing the scenario results in VEDA Back End by means of user defined sets, output

tables and graphs, and

Comparing and evaluating the scenarios to examine alternatives, develop insights, and

prepare recommendations relative to the initial analysis question.

The main activities associated with conducting a modelling analysis are summarised in Figure 1-1.

Chapter 3 presents some elements of the Demo model that is included in the software distribution

package. The text describes how to represent in a TIMES model the main components such as

energy and environment commodities, power plants, combined heat and power plants, mining or

importing / exporting an energy resource, and cars and trucks.

Figure 1-1 ETSAP tools and typical applications

[Acronyms: MARKAL = MARKet Allocation; TIMES = The Integrated MARKAL - EFOM System; LP = Linear

Programming; NLP = Non Linear Programming; VEDA = VErsatile Data Analyst user system; RD&D = Research,

Development & Deployment, ANSWER = user system]


11

2 How to model a simplified energy system: step-by-step description

This chapter introduces some base elements of the TIMES model generator and the VEDA users‟

interface. A simple example describes step-by-step how to get started with TIMES through the use

of VEDA templates6 and software. This simplified example is meant to familiarize you with the

main steps of the methodology and the main input, intermediate activities and output files.

2.1 A simplified energy system and its possible developments

2.1.1 Description of the example system and analysis questions

Your energy system consists of a single region (REG1). In the past (let‟s call it year 2000)

consumers demanded 15 EJ of Total Primary Energy Supply (TPES); in the future (let‟s assume

years 2001-2) the average demand is expected to grow to 16 EJ. In 2000 REG1 can produce

domestically 10 EJ/a of oil at the cost of 1 €/GJ; in 2001-2 an average of 9 EJ/a at the cost of 1.1

€/GJ can be produced. In fact you can increase domestic output of oil to say 12 EJ/a in 2001-2 if

you invest say in enhanced oil recovery from existing domestic fields or exploration of new oil

fields; this means that the extra oil extracted has a unit cost of 6 €/GJ. However, the domestic

extraction of oil produces associated CO2 emissions of about 10 KgCO2/GJ. Additional oil supply

can be imported at market prices7. In 2001-2 you can purchase GHG emission permits up to 20% of

2000 emissions at market prices. In order to use oil and satisfy their demand, consumers have to

buy a generic device (call it TECTPES) at the cost of 10 €/(GJ/a) and energy efficiency 1. In 2001-2

a new device is available at the cost of 12.75 €/(GJ/a) with an efficiency of 1.3 (the system data are

summarised in Table 2.1).

The task is:

- to understand how much your system costs, emits and depends on foreign supply without

policies;

- what emission reductions are possible; how they can be achieved and their trade-offs having

in mind that crude oil combustion emits about 70 kgCO2/GJ and purchasing GHG emission

permits costs always 10 €/tCO2eq;

- to what extent the dependence on foreign supply can be reduced, how this can be achieved

and its trade offs.

6 VEDA templates are (large) Excel workbooks in which the model data are assembled and managed by VEDA.

7 We learn from EIA (or IEA) statistics that crude oil import prices at the spot market have been 27.1 $/bbl in 2000 and

23.1 $/bbl in 2001-2 in the average (these prices are much lower than the futures). In the average 7 bbl of oil are

equivalent to 1 ton which has the energy content of 42.18 GJ. The inflation rate of USD with respect to the year 2000

has been of 2.9% in 2001 and 4.5% in 2002, which averages to 3.7%; the equivalent inflation of the euro has been

2.5%. The €-USD exchange rate has been 0.924 $/€ in 2000 and 0.921 $/€ in 2001-2 in the average (in July 2007 it

arrived to 1.38 $/€).


12

Table 2.1. Quantitative input data of the simplified energy supply sub-system

REG1 Unit Quantities (in unit) Cost (in M€2000/Unit)

2000 2001-2 ave. 2000 2001-2 ave.

Demand of TPES PJ 15000 16000

Domestic OILCRD production PJ 10000 9000/12000 1.0 1.1/6

Import of OILCRD PJ unbounded unbounded 4.90 4.04

Emission permits, purchase MtCO2 unbounded unbounded 5.0 6.0

Efficiency (decimal) Cost (in M€2000/unit)

Consumers’ device PJ/a 1.0 1.0/1.3 10 10/12.75

2.1.2

2.1.3 Description in TIMES: base elements

The main building blocks of a TIMES model are the processes and commodities, which are

connected by commodity flows in a network called a Reference Energy System (RES). The

dynamic part of a model is determined by the time horizon and resolution, the evolutionary

development of supply and technologies, the growth of the demand for energy services, and policies

(e.g., mitigation targets, renewable portfolio standards), complimented by various alternate

scenarios.

2.1.3.1 Commodities and processes

Commodities are assigned to different groups (Sets) according to their role in the energy system; in

this example the following sets are used: NRG = energy carriers, DEM = demands, ENV =

environmental indicators. The user must also specify the commodity unit, and remember that the

flows are measured in said commodity units. The commodities considered in this example are:

- total primary energy supply (TPES) is the demand commodity specified in units of PJ

(commodity type DEM);

- crude oil (OILCRD) is the supply commodity, specified in units of PJ (commodity type

NRG), and

- CO2 emission is an environmental commodity, specified in units of MtCO2 (commodity

type ENV).

Processes are similarly assigned to groups according to their role in the energy system; in this

example the following sets are used: IRE = inter-regional exchange/resource supply providing

energy to the region, PRE = energy processes converting energy from one form to another, and

DMD = demand devices that are a subset of PRE that consume energy to meet the demand for

energy services.

The processes used in this example, as shown in Figure 2.1, are:

- MINOILCRD1 (IRE) - represents the (limited) extraction of domestic crude oil at the cost of

1 €/GJ in 2000 and 1.1 €/GJ in 2001-2;


13

- MINOILCRD2 (IRE) - represents the potential (limited) extraction of domestic crude oil at

a higher cost in 2001-2;

- IMPOILCRD1 (IRE) - represents the (unlimited) import of crude oil from an external source

at the price of 4.9 €/GJ in 2000 and 4.1 €/GJ in 2000;

- TECTPES (PRE/DMD) - to transform the supply of OIL into the energy demanded by

consumers (TPES), and

- EXPCO2N - to represent the possibility to purchase emission permits8.

2.1.3.2 Possible system development: Scenarios

The development path of the system depends on the input assumptions and the objectives of the

policy makers. Both dimensions are explored through scenarios. The effect of different input

assumptions about the future development of system data is analysed running different “exploratory

scenarios” (omitted in this example). The effect of different policies is analysed running different

“control scenarios”. In this example the main control variables are:

Developing (or not) domestic oil,

Relying (or not) on oil import,

Improving (or not) energy efficiency of devices, and

Purchasing emission permits.

Three policy scenarios will be explored:

1. Least cost (base case)

2. Least energy dependence (independence), and

3. Least GHG emissions (mitigation).

Policy makers have to compromise among these objectives based upon their trade-offs calculated

by the analyst.

Total quantities and marginal prices of all commodities, as well as process capacities, activities and

the consumption/output levels for each associated commodity, are calculated by the model for each

scenario (see section 2.1.3.5). Starting from the model results, through simple calculations average

prices, carbon intensity of the system and other information are obtained.

2.1.3.3 Representation of the Reference Energy System

A TIMES model is structured by regions. One needs to distinguish between external regions and

internal regions. A TIMES model must consist of at least one internal region, and an external supply

region9. The internal region(s) correspond to regions within the area of study, and for which a

Reference Energy System (RES) is defined by the user. To handle domestic supply the designator

8 The typical best practice is to make an IRE take “imports” permits of -CO2 emissions at a price. Non standard option

adopted here illustrates the flexibility of the tool and the degree of freedoms of the user.

9 Using VEDA templates, it‟s necessary to declare the internal regions, while the external regions are automatically

generated by the software.


14

“MINRNW” is by convention used as an external region name, and for imports/exports

“IMPEXP”10

. Each internal region may contain processes and commodities to depict an energy

system, whereas external regions serve only as origins of commodities (e.g. for domestic supply of

primary energy resources (mining) or for the import of energy carriers) or as destination for the

export of commodities11

.

The RES for the simplified primary energy supply sub-system is shown in Figure 2-1. Imports and

mining come from an external region, pass through a process, and feed the final demand.

Figure 2-1 The Reference Energy System of the above simplified energy

10

To know more about the TIMES regions see the “Documentation for TIMES model – PART II” from page 16

available at http://www.etsap.org/Docs/TIMESDoc-Details.pdf. 11

This in necessary as the inter-regional exchange processes (IREs) used to define how commodities enter (domestic

supply/imports)/leave (exports) a region (e.g., trade) requires a from/to region designation.


15

2.1.3.4 The time dimension

In TIMES the time dimension is specified in terms of years and of division of a year.

The time horizon of the model is represented through model years. This problem encompasses three

years: 2000, 2001, 2002. The TIMES model represents this time horizon in a simplified way, by

periods: the first modelling period coincides with the year 2000 and has a one year length; the

second period represents the two years 2001-2, tagged by the former. The values associated with

this two year period are averages over 2001-2 (with the period total 2x that reported).

In order to combine costs, and in general economic variables, corresponding to different years in a

common numeraire, a discount rate in real terms (without inflation) has to be specified. In this case

a 7% discount rate is assumed, kept constant over the entire modelling horizon. The same rate is

used for depreciation of investments, unless a technology specific discount rate is provided.

The second time dimension of the model is sub-annual. For modelling electric load curves, space

heating demands and traffic, the year can be split according to seasonal, weekday / weekends, and

hours of the day characteristics, composed into time-slices. In this simplified case only an annual

level is modelled. In addition to the load imposed from the demand on the energy system, the

supply side may fluctuate within a year, e.g. hydro.

2.1.3.5 Calibration and Establishing the Reference Scenario

When building a model the first thing to address is the calibration of the 1st year of the model. This

entails configuring the resources and technologies specified to the model according to best

knowledge of what is in place in the initial model year, and ensuring that the model replicates this

situation. Once the model is calibrated the next essential step is to establish a Reference or business-

as-usual (BAU) scenario which describes the likely evolution of the energy system according to the

analysts‟ best judgement. The key aspects of the Reference scenario that need to be taken into

consideration are:

Determination of fuel energy prices and availability;

A projection of the future demand for energy services, based upon expected economic

growth (e.g., GDP) and other “drivers” (e.g., population);

Depiction of the retirement profile for existing power plants, upstream processes, and

demand devices, and know new builds;

A repository of future technology characterizations;

Controls to guide reasonable levels of fuel switching and technology penetration assuming a

nature evolution of the energy system, and

Any known policies to be imposed on the system (e.g., RPS, local pollutants and/or

greenhouse gas limits).

Once the Reference scenario is established the model can be used to examine alternative futures.


16

2.1.3.6 Scenarios

As already discussed, the Reference or BAU scenario describes the likely evolution of the energy

system, and serves as the starting point for conducting analyses using the model. Possible

alternative development paths of the system are explored by means of scenarios. Then these can be

compared with the BAU and other alternate scenarios to assess the implications of these alternative

futures, and formulate policy recommendations. One alternate scenario might look at the

implications of higher economic growth than was assumed for the BAU scenario. Another

alternative path might examine what is involved to stabilize CO2 emissions in 2001-2 at the same

level of 2000. A different point of view could lead to the maximisation of energy independence.

2.2 Declaration of the model through VEDA-FE12

2.2.1 Overview of VEDA-FE workbooks and internal syntax

VEDA-FE relies totally on templates, a collection of Excel workbooks, for all input data. The

VEDA templates permit the user to organise the information in flexible schema. The basic concepts

presented hereafter are intended to enable the user to wade through and digest the templates

associated with the TIMES_Tutorial model. Full details of the conventions, rules and options

available in the templates are described at www.kanors.com/vedasupport.

2.2.1.1 VEDA template controls

Inside the templates special characters “~” and “\I:” are reserved and indicate to VEDA-FE what to

do with the information that follows. The VEDA-FE import program reads each sheet in a

workbook in sequence, line-by-line from left to right. The basic types of codes (* is used here as a

wild card) are:

- Flexible Input codes (~FI_**:) establish the nature of the information to follow, they are

declared above a table and are valid for the whole table;

- Flexible Input Table (~FI_T) is the main data input indicator, and is placed in the row

immediately above the table headers and in the column before the first column containing

values [This identifier may not be followed by any comments or descriptions (use the \I:

ignore control to insert a comment on subsequent lines if desired). With this type of table

identifier, the data is imported as provided, that is it is not modified during the import

process (which may be case with the update control for example).];

- Ignore codes (\I: **) are declared in a table to ignore the rows and/or the columns where

they are specified, and

12

The VErsatile Data Analyst system is composed of two applications: the Front-End (FE) manages the input data and

submits model runs; and the Back-end (BE) handles the results.



17

- Special codes (~T**) are used to declare special tables, whose processing is different from

simple parameter insertion (e.g., ~TFM_FILL, a transformation directive to fill a table with

values from the base year templates (discussed below)).

2.2.1.2 VEDA rules

To define subsets of technologies in various types of VEDA templates (e.g., SubRES, Scenario,

UCs) a set of headers are used under which the user specifies masks using text/wildcards (“?” for a

single character, * for any number) to identify the qualifying technologies. Technology qualifiers

identify candidates based upon set members (Pset_Set), and/or masks for topology (commodity

in/out, Cset_CI/O), and/or name/description masks (Pset_PN/D). In general if nothing appears

below a specification column the values provided for the parameter apply to all entries. Exclude is

done by “-“<mask>. Multiple masks may be specified separated by “,”.

Although not explicitly shown in the templates, when data are read by VEDA and the internal

database is built, for every commodity a „dummy‟ import is generated – whose code usually ends

with Z. The amounts are not bounded but the cost is a few orders of magnitude higher than real

import costs.

Some of the parameters described in the guide book and supported from VEDA, must be combined

with the limtype (UP, LO and FX), see in VEDA-FE interface Tools, Supported Attributes; and

optionally a year (the value applies to base year when not specified). Therefore,

BNDACT~2001~UP is an upper bound on resource technology activity for the year 2001. Instead

BNDACT~UP is an upper bound on resource technology activity for the base year (in this example

2000).

2.2.1.3 VEDA workbooks

Templates are Excel spreadsheet workbooks that lay down the basic structure of the model and hold

the fundamental data and assumptions. These templates provide the information about the base

year, demand projection(s), future technology possibilities and scenario assumptions that

collectively describe the entire energy system to be studied13

. The templates are14

:

- SysSettings: this is one of the three files in VEDA with a fixed name, which stands for

System Settings. It is used to declare the very basic model structure like regions, timeslices,

start year etc. It also contains interpolation rules for various attributes and some settings for

the synchronization process.

13

In many cases VEDA-FE provides defaults that often meet the user‟s needs and leave the user to focus of the actual

data. 14

In this example, the templates are built by hand. Often data are fed to the base templates from existing statistical

spreadsheets and databases, perhaps with a direct linkage to existing data source (e.g., IEA and EUROSTAT energy

balance statistics).


18

- Base Year Templates: these are meant to set up the base-year technology stock and demand

levels such that the overall energy flows respect the energy balance. In other words, start

year of the model is calibrated in the B-Y templates.

- The B-Y templates are named as VT_<super region>_<sector>_<Version>15

. The first two

“_” are crucial for VEDA to interpret the file as a valid B-Y template.

In these files, we create processes, commodities, create the RES connections (commodity

input/output of processes) and add the basic attributes for processes and commodities. By

default, all declarations hold for all the regions that map to the super region indicated by the

file name. But all the above declarations can be made region-specific by introducing a

region column in any of the tables. The process characterization table supports~<Region

Name> in the column headers.

- BY_Trans_<anything>: this file supports all tables and works just like a scenario file, with

one important difference: the process and commodity filters (rule-based16

) see only those

elements that come from the B-Y templates. One can have alternate versions of this file

and select one at the time of SYNC operation.

- SubRes_<application> Data/Transformation files: new technology (and sub-RES)

definitions data specification and regional transformation (for the respective SubRES) (rule-

based); Declarations in SubRES files are identical to the B-Y templates with one important

difference: all are completely region independent. Region- specificity is introduced via the

process availability table17

and transformation tables in the related transformation file.

- Demand Module: for demand drivers (economic indicators), and elasticities series used in

the region/segment driver allocation table;

- Scen_<scenario designator> Scenarios files: contains data specifications and transformation

for any part of the RES (rule-based);

- ScenTrade_<trade scenario> Trade scenarios: parameters specification for the trade

technologies, which is the only way to create parameters for the IRE processes (rule-based),

and

- ScenUC_<constraint scenario> UC Scenarios: the special share constraints; based upon the

initial fuel shares in the base year templates, with user provided transformations applied to

control the rate at which these shares may or may not change over time.

2.2.2 Sub-folders structure

VEDA components are organised in a specific folder structure. The main folder for the models

managed by VEDA is \VEDA\VEDA_Models, with a sub-folder for each individual models a user

is working with (see Figure 2-2).

15

For example for the supply sector should be VT_REG1_SUP_V1p0.xls; VT designates this Excel workbook as a

VEDA-Template for Region1, Supply sector, and version 1p0. 16

“Rule-based” means that groups of technologies may be qualified and operations then applied to data based upon

name/description character masks, input/output commodities (topology), and set membership. 17

For more information see VEDA-FE/Sub-Res Transformation file at www.kanors.com/vedasupport.



19

Figure 2-2 VEDA_Models Folder

The TIMES tutorial example is stored in the folder TIMES_TUTORIAL (TT). In this folder there

are three sub-folders – Databases, SubRES_TMPL, and SuppXLS – and three templates –

SysSettings (with the region, time-slice, periods et al declarations, as interpolation/extrapolation

parameter defaults), BY_Trans and VT_TT_SUP_V1p0 (with the base year data, full-blown models

will likely have several base year template files) (see Figure 2-3).

Figure 2-3 Folder TIMES Tutorial

The Databases sub-folders is maintained by VEDA-FE, the SubRES_TMPL folder contains two

workbooks as shown in Figure 2-4 containing the future technology options, and the SuppXLS

folder contains three workbooks as shown in Figure 2-5. where alternate scenarios are kept. In the

next sections details regarding how to construct the two template files (SysSettings and

VT_TT_SUP_V1p0) are discussed.


20

Figure 2-4 Sub-folder SubRES_TMPL

Figure 2-5 Sub-folder SuppXLS

2.2.3 SysSettings: Model Setup Template

The SysSettings file contains comprehensive information about the model‟s basic structure (regions,

sub-regions, time-periods and time-slices), along with the default interpolation and extrapolation

user input definition. This workbook is made up of various sheets, listed here and then elaborated

on in the follow three sections:

1. RTT (see Figure 2-6) - for the regions, time periods and time-slices definition;

2. Interpolation. - for default interpolation rules and parameter values, and

3. Constants – for discount rate and fraction of year for time slice.

4. Create CG - for commodity group definitions.


21

2.2.3.1 Declaring regions and time-slices

The first step is to define the regions and time-slices. Start by opening the “SysSettings” workbook.

Figure 2-6 SysSettings: Example 1 – Regions and Time slices related declarations

The two tables shown above need to appear on a sheet called Region-Time Slices. There are no

other restrictions on sheet names in this file. In this sheet to define the regions and time-slices

“Workbook Controls” are used as shown in Row-1 above, and explained below.

- The workbook control ~BookRegions_Map (cell A1) is used to define:

o the template root (cell A3,4,5,..) for the base year sector template(s): TT (TIMES

Tutorial); the name must be the same for each base year template workbook

VT_TT_<sector>_<version> (in this case just 1 base year templates is involved

VT_TT_SUP_V1p0), and

o the list of region names (cell B3,4,5,...) maps the names used in the B-Y templates to

the model regions: in this case just REG1.

- The workbook control ~TimeSlices (cell E1) is used to define:

o the time-slices resolution for the model, by declaring the elements of each time-slice.

In the column-E SEASON, in column-F WEEKLY in column-G DAYNITE and in

column-K ANNUAL. In this example there is only one time-slice (ANNUAL), as

shown in Figure 2-6.

2.2.3.2 Declaring model horizon

The second step is to define the model horizon in the “SysSettings” workbook, sheet TimePeriods.

Figure 2-7 SysSettings: Example 2 – Model horizon declarations


22

In this sheet to define the start year and time periods “Workbook Controls” are used as shown in the

Figure above, and explained below.

- The workbook control ~StartYear (cell B3) is used to define:

o the start year of the model. In the cell B4 is the base year (in this example 2000).

- The workbook control ~ActivePDef (cell B8) is used to define:

o the set of active periods. Alternate period definitions can be made and the active one

is declared under the tag ~ActiveDef. For example, PeriodsDef (cell B8) is the active

set periods for the TT.

- The workbook control ~TimePeriods (cell B11) is used to define:

o the time horizon of the model for the ActivePdef. In cell B13 the number of years for

the start year (first period is 1 year long), the second period has 2 years.

2.2.3.3 Declaring settings for the synchronization process

This sheet contains some settings for the synchronization process.

Figure 2-8 SysSettings: Import Settings

- Dummy Imports18

: one can control the creation of dummy imports in the table shown above.

- Vintage Bounds: VEDA can automatically generate bounds to prohibit investments in

processes when a newer vintage becomes available, assuming that the last two characters of

the process name are used for the first year of availability. For example, TCAR05 will no

longer be available for investment after TCAR15 becomes available. This feature is one of

the surviving relics from the MARKAL days of VEDA, when full vintaging was not

available. It can still be used as a parsimonious substitute for full vintaging option of

TIMES. However, users must be much disciplined in naming processes while using this

option.

All tables described so far are specific to the SysSettings file and are not supported in any other

file in the system. The tables above are read only when the date-time stamp of this file changes.

18

These backstop dummy processes are introduced in the model in order to avoid infeasibilities that may arise if not

enough energy carrier or demand can be supplied.


23

2.2.3.4 Declaring inter/extrapolation options and dummy import prices

Over and above all the special tables, SysSettings supports all the tables used in regular scenario

files. It is recommended to have the following declarations via INS and UPD tables shown below:

o Interpolation settings

o Dummy Import prices

o Global constants – discount rate and year fraction in each timeslice

The next step is to designate the defaults interpolations and extrapolations options. In the

Interpol_Extrapol_Defaults sheet of the file SysSettings (see Figure 2-9) users declare the rules for

inter/extrapolating input parameters that are time dependent, and other data manipulation options19

.

To define the inter/extrapolations options, the “Workbook Controls” discussed below are used.

- The workbook control ~TFM_UPD is used to declare a table in a scenario file which is a

simple transformation to pre-existing data in a rule-based manner. In this case, the 1st block

deals with setting the default interpolation rules (indicated by the 0 in the Yr column) for

various parameters20

. The 2nd

block sets default prices (ACTCOST) for the “backstop”

options for fuels (dummy IMPort technologies ending with “Z”) and demands (a dummy

IMPDEMZ process that can feed any demand).

- The data values are normally preceded by a mathematical operator, like *.5, +4 etc. *, +, -

and / are supported as the first character.

- These tables support absolute values as well. By default, these tables are read as text. If one

wants to have only absolute values in an update table, it is recommended to use the tag

“~TFM_UPD-N”, to force the table to be read as numeric.

All column headers and five fields to identify processes and three to identify commodities are

described in "Templates Basic - tables" at www.kanors.com/vedasupport.

Figure 2-9 SysSettings – Inter_Extrapolation and Default options sheet

19

The rules are fully described in the Appendix III, paragraph Error! Reference source not found.. Error! Reference

source not found.. 20

See the “Documentation for the TIMES model – Part II – page 39” for a full description of the (powerful)

interpolation/extrapolation facility in TIMES.



24

- The price of the backstop options is set, based upon the technology name masks provided.

Note that the IMPDEMZ specification calls for 4x the existing value, which is set to 2222

on the line above.

2.2.3.5 Declaring Constants

This sheet is used to define some overall parameters to the model. The only workbook control used

in this sheet is ~TFM_INS (transformation insert table21

), to assign an absolute value for

parameters that are based on rules.

- ATTRIBUTE, a reserved word indicating that a VEDA-TIMES parameter names appear in

this column; in this example these parameters are DISCOUNT (overall discount rate) and

YRFR (fraction of year for season, time-of-day; = 1 because there is only the annual time-

slice.), and

- TIMESLICE indicating the particular time-slice for which data is provided, where in this

case that there is only an annual time-slice22

.

Figure 2-10 SysSettings - Constants sheet

2.2.4 Commodities and processes definition in the base year template file

VT_TT_SUP_V1p0

In general, to define a commodity in TIMES-VEDA the user must specify:

- commodity set membership;

- commodity name and description;

- commodity unit;

- the sense of the Balance equation;

- timeslice level

- peak monitoring

- commodity group (other than “own commodity groups” and the defaults created by VEDA

based upon the nature of a process, as just discussed);

- commodity type (it is a flag for electricity commodity).

To define a basic process in TIMES-VEDA the user must specify:

- process set membership;

- process region;

21

For more information see Transformation tables, Insert Value section at www.kanors.com/vedasupport. 22

The full description of the time-slices is given in paragraph 3.2.1.14.2.3.2.


25

- process name and description;

- process unit (activity and capacity)

- timeslice level of process activity

- primary commodity (group) upon which process operation is defined (if not the VEDA

default of output consisting all the commodities of the type according to the nature of the

process), and

- the tracking vintage;

The following paragraphs describe how this information is declared in the base year template.

Note that the approach described here is meant to illustrate good practice, but it is up to each user to

decide how they want to organize their templates and construct their model. Furthermore, the names

used for each of the sheets discussed here are also user defined. However the various VEDA

controls, column headers, and parameters names are pre-defined and must be used appropriately.

In general the supply template (SUP) describes fossil fuel extraction, renewable potentials, and

various fuel transformation processes including petroleum refineries and gas pipelines. For the

Tutorial model this is done in a single file, VT_TT_SUP_V1p0, but as already mentioned more

robust models may have multiple base year templates, one for each sector. In this example in the

SUP template describes the crude oil extraction and import, and the total primary energy supply

demand. In this particular file there are seven sheets, each discussed in the subsequent sections.

2.2.4.1 Balance Sheet

The Balance sheet contains the fundamental data about the base year energy balance. In this very

simplified model it simply shows the import and domestic production of crude oil and the associate

total consumption of same. The Balance sheet as such is not needed to by the model; therefore it

does not contain VEDA control characters and it is neither read nor imported into VEDA-FE (see

Figure 2-11); the energy balance sheet is used to calibrate – in other sheets – quantities of

commodities and technologies in the base year, as described below.

Figure 2-11 Balance sheet of VT_TT_SUP_V1p0


26

2.2.4.2 SUP_Comm Sheet

The SUP_Comm sheet identifies the individual commodities found in the workbook. The

commodity type indicates the nature of a commodity (energy (NRG), material (MAT), demand

service (DEM), emissions (ENV) and financial (FIN)). It determines as well the default type of

constraint of the commodity balance equation: for NRG, ENV and DEM the supply has to be

greater than (default), or equal to, consumption, while for MAT and FIN equality holds.

In VEDA templates all commodities must be declared and the type defined, as shown in the

SUP_Comm sheet (see Figure 2-12).

Figure 2-12 SUP_Comm sheet of VT_TT_SUP_V1p0

In Figure 2-12 the workbook control ~FI_Comm is used at the upper left corner of the specification

area to declare the following column headers:

- Csets: the sets to which commodities belong; the commodity sets indicates the nature of a

commodity (energy (NRG), material (MAT), demand service (DEM), emissions (ENV)

and financial (FIN)). Csets declaration are inherited until the next one is encountered.

OILCRD is an energy (NRG) commodity ( (thus if the model also used hard coal

(COAHAR), and natural gas (GASNAT) these could follow OILCRD without Csets entries

and would be designated as NRG commodities)), TPES is a demand commodity (DEM)

and CO2 an environmental (ENV) commodity;

- CommName: commodity name, which needs to be unique;

- CommDesc: commodity description;

- Unit: the commodity unit throughout the model (the user should note that within the TIMES

GAMS code no unit conversion, therefore, the proper handling of the units is entirely the

responsibility of the user);

- LimType: the sense of the balance equation (which may be LO (Production>=Consumption,

FX (Production=Consumption), UP (Production<=Consumption), NON (no-binding,

accounting only ) allows for the user to adjust the default equation type, though this usually

not necessary; CSets determines the default type of constraint of the commodity balance

equation: for NRG, ENV and DEM is LO while for MAT and FIN is FX.


27

- CTSLvl: the commodity time-slice tracking level; valid entries are ANNUAL, SEASON,

WEEKLY and DAYNITE (defaults to ANNUAL when not specified);

- PeakTS: peak timeslice monitoring; it is allowed ANNUAL or user timeslices defined in

the SysSettings. ANNUAL generates the peaking equation for all timeslices and user

timeslices generates peaking equation only for the defined TS in the table (Comma-

separated entries allowed;

- CType: electricity commodities indicator (ELC).

Note that the Commodity definition sheet requires a fixed layout of the column headers, as shown

here, which may not be changed.

2.2.4.3 SUP_Process Sheet

The SUP_Process sheet identifies the individual processes found in the workbook. In VEDA

templates all processes must be declared and the type defined, as shown in the SUP_Process sheet

(see ).

Figure 2-13 SUP_Process sheet of VT_TT_SUP_V1p0

In Figure 2-13 the workbook control ~FI_Process is used at the upper left corner of the

specification area to declare the following column headers:

- Sets: the sets to which processes belong; the process sets indicates the nature of a process

(thermal electric power plant (ELE), combined heat and power (CHP), heating plant (HPL),

pump storage (STGTSS), pump storage IP (STGIPS), generic process/technology (PRE),

demand device (DMD), import process (IMP), export process (EXP), mining process (MIN)

and renewable potential technology (RNW).

- Region: region(s) in which the processes exist; comma-separated entries are allowed. If the

field is blank or omitted then the declaration applies to all regions related to the workbook.

Declarations are not inherited. This Column is used only for BY Templates

- TechName: name of the process, which needs to be unique; no spaces; up to 32 characters

long


28

- ProcessDesc: description of the process; up to 255 characters long;

- Tact: activity unit of the process;

- Tcap: capacity unit of the process;

- Tslvl: the process time-slice operational level; valid entries are ANNUAL, SEASON,

WEEKLY and DAYNITE. If empty default based on Sets declaration (ELE, STGTSS and

STGIPS=DAYNITE, CHP and HPL=SEASON and other processes ANNUAL)

- PrimaryCG: Primary Commodity Group of the process. This declaration is needed only

when one wants to override the default PCG allocated by VEDA23

.

- Vintage: electricity Vintage tracking. Only YES/NO entries are allowed (if empty default is

NO).

2.2.5 Process characterization24

file VT_TT_SUP_V1p0

Topology (input and output of processes) and parameters for technologies and commodities are

defined using tables as shown in Figure 2-14. While the parameter information can be injected via

several other tables, this is the only way to declare the topology information. As the indexes for

numerical entries can be specified as row identifiers or column headers, this is by far the most

flexible table in VEDA. It can be configured in a variety of different ways depending on the

requirement.

Figure 2-14 Topology and parameter defnition table

In the next paragraphs some the column headers and some of the possible entries are described for

the VT_TT_SUP_v1p0 template. For more information see Templates Basic Tables, Process

Characterization ar www.kanors.com/vedasupport.

23

For more information see paragraph 3.2.1.2. 24

For more information see Templates Basic Tables. Process Characterization at www.kanors.com/vedasupport.



29

2.2.5.1 MIN Sheet

This sheet is used to characterize the domestic resource supply options for crude oil. In this sheet, to

characteriza the supply processes, the “~FI_T” “Workbook Control” is used:

- “~FI_T” is used to declare a data table.

The ID column headers used to characterize the resource supply technologies of the Tutorial model

are described here25

and shown in Figure 2-15.

- TechName - defines the technology name. The technology name must be unique, though

the same techname can be used in different regions.

- TechDesc - describes the technology.

- Comm-Out - identifies the name of the output commodity from a process (note that

resource supply options are (usually) to the processes that do not have an input).

Figure 2-15 MIN sheet of VT_TT_SUP_V1p0

The Data Area Column headers used to characterize the resource supply technologies of the

Tutorial model are described here26

.

- CUM - declares the cumulative level of a resource over the modelling horizon, in user

defined units.

- COST - assigns the annual resource cost per unit of production/external trade, in user

defined units (market price/energy unit).

- BNDACT - specifies a bound on the annual activity of a (resource) technology. It must be

combined with the limtype (UP, LO and FX); and optionally a year (the value applies to

base year when not specified). Therefore, BNDACT~2001~UP is an upper bound on

resource technology activity for the year 2001.

- ENV_ACT specifies the value (emission rate = unit emission/unit of commodity produced)

of an emission declared in the Comm-OUT column.

25

All permitted ID column headers for the templates are described in Appendix paragraph Error! Reference source

not found., Error! Reference source not found.. 26

All permitted Area Dara column headers for the templates are described in VEDA user interface under

Tools/Supported Attributes.


30

Note that the units are provided as informative comments on a "*" row 6.

With regard to the data, the user provides the technology parameters in the grey columns, such as

cumulative reserves, costs and/or bounds. In this example there is a one step supply curve for 2000

and a two step supply curve for the period 2001-2.

In general base year templates hold the structure and assumptions about energy production and

consumption for the base year. They provide the information related to the base year energy

balances and the technology stock. In each sector, the energy production and consumption are

calibrated in the templates to match base-year statistics for the initial time-period.

2.2.5.2 IMP-EXP Sheet

The IMP-EXP sheet contains the technology information (costs and bounds) related to the

exogenous energy trade for crude oil (OILCRD). In this sheet the “~FI_T”: “Workbook Control” is

used to characterize the exogenous process:

Figure 2-16 IMP-EXP sheet of VT_TT_SUP_V1p0 – IMPOILCRD1

In this sheet there is also information about another technology used to represent the possibility to

purchase emission permits. This technology, shown in Figure 2-17, contains information about the

cost associated with the emission permits of the commodity CO2N.

In this case in the cells F18 and G18 there is a negative cost, used to simulate the purchase of the

emission permits27

.

27

The typical best practice is to make an IRE take “imports” permits of -CO2 emissions at a price. Non standard option

adopted here illustrates the flexibility of the tool and the degree of freedoms of the user.


31

Figure 2-17 IMP-EXP sheet of VT_TT_SUP_V1p0 – Emission Permits CO2N

2.2.5.3 DEMAND Sheet

This sheet is used to characterize the demand technologies and service demand levels for the base

year. In this sheet, to define the demand process and demand value the “Workbook Control”

“~FI_T” is used :

This worksheet is split into two parts, the 1st providing the details on the demand technologies, and

the 2nd

the demand themselves.

The ID column headers used to characterize the demand technologies are described here and shown

in Figure 2-18.

- TechName - defines the name of the demand devices;

- TechDesc - describes the devices;

- Comm-IN - indicates the name of the commodities input to a process, in this case the

process consumes oil (OILCRD);

Comm-OUT – indicates the name of the commodities output from a process, in this case the

process produces demand services (TPES);

The Data Area column headers used to characterize the demand technologies are described here and

shown in Figure 2-18.

- EFF - specifies the demand technology efficiency (sum outputs/sum inputs);

- INVCOST - assigns the demand technology investment cost (market price/capacity unit) in

user defined units, and

- LIFE - assigns the demand technology life in years.


32

Note that in this example investments are permitted in the existing technologies, as there is no

specification by the user limiting this possibility. In general in the calibration templates it is

preferable to prevent investment in existing technologies (by means of IBOND-UP = 0), because

the purpose of the base year templates is to calibrate the energy flow and the existing stock for the

base year energy balance.

With respect to the demands, a new ~FI_T declares a separate table in which there are the

following column headers:

- CommName - identifies the demand itself (ID column header);

- DEMAND~2000 - defines the demand level for 2000 (data area column header);

- DEMAND~2001 - defines the demand level for 2001 and 2002 (data area column header).

Figure 2-18 Demand sheet of VT_TT_SUP_V1p0

2.2.5.4 EMIssion Sheet

This sheet contains the emission factors. The workbook control used on this sheet is ~COMEMI

(to link emissions to consumption of a commodity), to declare a special table, whose processing is

different from the standard flexible input table designator, ~FI_T. With this type of table identifier

the data are manipulated during the import process not simply imported as provided.

The emission factor provided for CO2N (CO2 produced by combustion), via ~COMEMI, will be

associated with each of the commodities listed as column headers for which an emission coefficient

is provided. Thus in this case the emissions are charged when OILCRD is consumed, by TECTEPS

(where 70 kt/PJ is the approximate CO2 emission factor of crude oil).


33

Figure 2-19 Emission sheet of VT_TT_SUP_V1p0

Fugitive CO2 emissions in the oil extraction phase are associated to processes MINOILCRD1/2 and

are declared in the MIN sheet (see above, paragraph ).

2.2.6 SuppXlS template folder

In general the SuppXlS folder includes Scenario files, with data specifications and transformation

for the entire RES, and two sub folders (Demands and Trades). In this case there are two scenario

files and two sub-folders, as shown in Figure 2-20. Note that in this example some of the folders

and files may be empty, however they will be explained in subsequent examples.

Figure 2-20 SuppXlS folder


34

2.2.6.1 Scenario files

The Scen_<appl>.xls files contain additional information and parameters for the entire RES,

commodities and technologies (rule-based). The important thing to understand about scenario files

is that they can only manipulate information associated with previously declared RES components,

and that new commodities and technologies may not be added via scenario files, though parameters

may be. For this example there are two files:

- Scen_CO2N_Bound, containing the emission cap used for the scenario which forces

stabilization of CO2N emission. Figure 2-21 shows the parameter COM_BNDNET used to

give a bound on CO2 emission28

. This is an annual upper emission bound (LimType-UP) of

1150000 (Kt) for the environmental commodity (ENV) CO2N.

Figure 2-21 Scen_CO2N_Bound

- Scen_IMP_Bound, used for the scenario looking at minimum imports. Figure 2-22 shows

the parameter ACT_BND used to give an activity bound29. This is an annual lower bound

(4500 PJ) on the activity of the technology IMPOILCRD1 for the year 2001.

Figure 2-22 Scen_IMP_Bound

2.2.6.2 Demand sub-folder

For this example the folder is empty. It would normally contain the demand drivers and their

relationship to the individual sector demands. However, since the demand was explicitly specified

directly in the calibration template (for both 2000/2001) no additional information is required.

2.2.6.3 Trades sub-folder

For this example the folder is empty, since this is a single region model. Normally a file would

establish the traded commodities between the internal regions (which may be established manually

or via the VEDA-FE Advanced/Trade facility (recommended)). For the trade matrix intersections

28

The scenario mask is described in the paragraph 2.2.1.2. 29

The scenario mask is described in the paragraph 2.2.1.2.


35

VEDA-FE constructs appropriate TIMES IRE (inter-regional trade) processes to which the

parameters may then be associated. Any number of trade scenario files with data to be associated

with the permitted trade links (IRE processes) may also be provided (e.g., bounds, investment costs

for expanding electricity grids/pipelines, additional transportation costs).

2.2.6.4 UConstraints sub-folder

For this example the folder is empty. It would normally contain the user define ad hoc constraints

that maybe introduced to the model (discussed in Chapter 3).

2.2.7 SubRES_TMPL templates folder

The SubRES_TMPL folder contains the technology declarations/data and mapping workbooks for

various subsets of the energy system (including the entire RES). The files associated with defining

SubRES and assigning them to regions is most often used for multi-region models so that a single

declaration can be made for all regions, and then just the variations in said declarations applied

where necessary (e.g., if the efficiency of a coal plant is region X is characteristically 90% of the

norm, then a transformation on the EFF parameter of *0.9 would be applied, according to a

Pset_PD = *COAL* and Pset_Set = ELE, assuming that all such plants have Coal in their

descriptions).

The files in this folder are all to named SubRES<user-name>_Trans, where it is required to have the

pair for each <user-name> SubRES scenario.

2.2.7.1 SubRES-B-NewTechs

The SubRes Data files contain the new technology definitions and data specification (rule-based).

For this example the file is empty, as all technologies have been defined in the base year templates.

2.2.7.2 SubRES-B-NewTechs_Trans

The SubRes Transformation files contain the mapping and transformation operations that control

the inheritance (or not) of new technology into the various regions. For this example the file is

empty, as all technologies have been defined in the base year templates.


36

2.3 Managing the Templates via VEDA-FE

In order to generate a TIMES model, the information declared in the templates is gathered into a

single database. This section describes how to import the templates into VEDA-FE and check that

the data in the database correspond to the intended specifications. The next section describes how to

run the simplified energy supply sub-model. Both steps are achieved by launching the VEDA-FE

program (double-click on the “open book” icon created on the desktop).

2.3.1 The VEDA-FE Navigator

VEDA-FE opens displaying the VEDA-Navigator, which provides a comprehensive view of all the

templates in the various folders managed by VEDA for the current model. It is the main vehicle for

accessing, importing and coordinating the various templates. As can be seen in Figure 2-23 the

VEDA-Navigator front screen is divided into sub-windows according to the various types of

templates managed by VEDA as discussed in the previous sections.

Specifically, the groups are:

- B-Y (Base Year) Templates;

- BY_Trans;

- Scenario Files;

- SubRES Data Files;

- Demand Files;

- Trade modules, and

- User Constraints.

Through this straightforward organization VEDA-Navigator enables easy access to any of the Excel

workbooks constituting the current model being used. To do so the user may either click directly on

a template name or using the bar above each area open Windows Explore in the associated folder

(for example by clicking on SubRes Data File, the folder that will be open is

C:\VEDA\VEDA_Models\TIMES_TUTORIAL\SubRes).

The VEDA-Navigator also provides feedback as to status of the various templates and the actual

assembled database managed by VEDA. This can be seen by the coloration of the individual

templates as well as the database icons according to the legend at the bottom of the form, discussed

subsequently.

2.3.2 Select the model to be processed

The current active model appears on the bottom right line of the VEDA-FE window (not shown),

with the associated template folder displayed on the top line of the VEDA-Navigator form. When


37

you initially install VEDA it is configured for the Demo Model, and the VEDA-Navigator window

will be similar to that shown in

To select a different model double-click on the MODEL: Template Folder bar. A new window is

shown in Figure 2-24 (Select Template Path), which will display the current list of VEDA model

folders associated with your installation. In this window click „NEW‟ and navigate to the path of

the template folder for the TIMES Tutorial example

(C:\VEDA\VEDA_Models\TIMES_TUTORIAL), then highlight the new path and click „OK‟.

Error! Reference source not found.Figure 2-23 The VEDA-FE Navigator

Figure 2-24 Active models in VEDA-Navigator

2.3.3 Synchronize model templates and internal database

Figure 2-25 shows the templates of the TIMES Tutorial model in the VEDA-Navigator. The

consistency of the templates and the integrated database managed by VEDA is immediately evident

based upon whether the ALL OK (yellow # 4) icon is shaded (red) or not (if any template has a

newer date/time stamp than the databases then an inconsistent state exists). In addition, the user can

see the base year template in red (yellow label 2) indicating that the template is inconsistent, and

that the B-NewTechs has not yet been imported owing to being displayed in white.

The way to synchronize the templates in the application folder is to hit the SYNC button. The files

that is imported is the red one (white (not imported), light blue (consistent), yellow (to delete) and

black (missing))).

To import a template that has not yet been brought into VEDA it is enough to select the template, in

this case SubRes (yellow label-3) and Scenario files (yellow label-5), by checking it and then click

„SYNC‟ (yellow label-4). After this you will see templates processing and at the end of that process

the base year template and the SubRes file became consistent (see

Figure 2-27).


38

Figure 2-25 The VEDA Front End Navigator Figure 2-26 Processing templates

Figure 2-27 Templates and databases are consistent


39

2.3.4 Browse/Edit data in VEDA-FE database

Once the templates have been imported and assembled as a database within VEDA-FE it is possible

to review the resulting data by means of powerful filtering tools and dynamic data cubes (pivot

tables), and it is also possible to check the RES by requesting the network diagram be displayed.

2.3.4.1 Search and view data

In order to view (and directly edit) the data in VEDA-FE, the user needs to follow these steps.

1) From the starting window, select „Basic Functions‟, „Browse/Edit‟ and then „TIMES View

or VEDA View‟ (alternatively [F7 or Shift-F7]), as shown in Figure 2-50. The TIMES View

option, will show you the TIMES parameter, instead the VEDA View, which shows the

same information but with the parameters names used in the templates30

. In Figure 2-29 is

shown the TIMES View Browse.

Figure 2-28 VEDA-FE Browse/Edit

30

If in TIMES view you have the ACT_BND, in VEDA view there will be the same parameter called BNDACT (see

VEDA-FE interface under Tools, Supported Attributes).


40

2) Using the TIMES View (or the VEDA View) form (see Figure 2-29) the user can view the

subsets of the assembled data in a cube by selecting the scenario(s), region(s), the

process(es), the commodity(ies), and/or the attribute(s) of interest.

Figure 2-29 VEDA-FE TIMES View Browse

To view all the data related to the REG1, check REG1 (yellow label 1) and then simply click

the „Browse‟ button (yellow label 2). A new window is shown in Figure 2-30 with a

dynamic data cube presenting all the associated information.

Figure 2-30 Example of dynamic data cube (pivot table)


41

The user can rearrange the layout of the cube by adding/removing dimensions (columns and

rows) to the table (as shown in Figure 2-31 in which the process is unselected as outermost

row label) by dragging and dropping components from/to the area above the current row

designator columns. When the cursor is placed over a dimension a crosshair appears. Then,

holding the left mouse button down and sliding to a new position a green line will appear

indicating that the dimension may now be “dropped” there. Any dimension not positioned as

part of the row/column table layout definition appears at the top of the page. These

dimensions have their values summed in the cube. For each dimension on top of the page if

more than 1 value exists for that dimension its name will be displayed between two asterisks

reinforcing that some values in the cube may be aggregates. [Note that for any dimension

where only a single value exists said dimension is automatically moved up top.] With the

pull-down arrow associated with each header individual entries may be temporarily removed

by unselecting it from the list of elements.

Figure 2-31 Example of dynamic data cube rearrange

3) A filter mechanism (VEDA-FE Search Engine) helps the user to quickly select only the

subset of the RES and associated data of interest by selecting the interested component by

clicking on an element (e.g. Process) in the associate list-box and then press F3 or use the

right-mouse and select Search, to see the new window in Figure 2-32. The filter mechanism

works by:

- string matching based upon start/any/exact, without wildcards;

- using concatenated criteria separated by „,‟;

- via short name and/or description;

- using include/excluded based upon above, and

- incrementally adding to existing selection list, or 1st clearing the current selections.


42

For example, to view only the data related to the MINOILCRD1, the user can select

„Starting With‟, write the short name MINO in the box „Include elements where‟ and write 2

in the short name in the box „Exclude elements where‟, Select All (which checks the process

in the list) and then Browse. The new window is shown in Figure 2-33.

As already noted, the user can then rearrange the layout of the cube as desired by dragging

and dropping components from/to the area above the current row designator columns (in

Figure 2-34 the process is unselected as outermost row label by moving it up top as only 1

process is involved).

Figure 2-32 VEDA-FE Search Engine - Processes

Figure 2-33 Example of filter use in VEDA-FE browse


43

4) To edit the data from the browse screen, the user may click direct edit (yellow label 1) and

than double click on the data to be edited, as shown in Figure 2-34 (see yellow label 2).

Figure 2-34 Example of direct edit data from VEDA-FE browse

VEDA-FE then opens a pop-up window (see Figure 2-35) in which it is possible to modify

directly the data. For example the user can try to change the ACT_BND from 9000 to

10000, then click OK to get to Figure 2-36. Note that VEDA-FE has actually updated the

associated spreadsheet, retaining the previous value as a comment in the associated cell, if

requested on the confirmation form.

Figure 2-35 How to change a data from direct edit

Figure 2-36 FE Navigator after the direct edit

It is possible to check the update made with the direct edit in the base year templates, as well

as other templates not involving transformations of data. The user can get back to the

Navigator from the „Basic Functions‟ menu, or by pressing F6 function key (see Figure


44

2-37), or arrange the Browse window as an overlay and simply move back and forth by

clicking on the forms. Then double click on the file VT_TT_SUP_V1p0 (yellow label in

Figure 2-37) to open the base year template and check the update (see Figure 2-38).

If one wants to obtain the same results shown in the next paragraphs, one has to change the

new value 10000 back to 9000, and follow the previous steps to update the VEDA-FE

database.

Figure 2-37 FE Navigator after the edit

Figure 2-38 The base year template updated


45

2.3.4.2 Process and Commodity Master

The process master (see Figure 2-39) show detailed process information like set memberships, units

and input/output and in the data tab shows all associated parameter information. From the process

master view is shown the set membership of input/output commodities in color-coded and double-

clicking a commodity opens the commodity master for that commodity.

Figure 2-39 Process Master

The commodity master (see Figure 2-40) show detailed commodity information like set

memberships, units and producers/consumers and in the data tab shows all associated parameter

information. From the commodity master view is shown the set membership of

producing/consuming processes in color-coded and double-clicking a process opens the commodity

master for that process.

Figure 2-40 Commodity Master


46

2.3.4.3 Commodity Group Master

The commodity group master (see Figure 2-41) shows the commodities associated with all user-

defined and VEDA-created commodity groups and the set membership of input/output commodites

in color-coded. Double clicking a member commodity opens the commodity master for that

commodity.

Figure 2-41 Commodity Group Master

2.3.4.4 RES view

In order to view the RES, the user needs to select „Basic Functions‟ and then „RES,‟ or press the F8

function key, which presents something like Figure 2-42, depending upon what information is

currently selected in the Browse. In this window it is possible to navigate around the model by

clicking on the name of the commodities or technologies, which allows the user to see in case of a

commodity all producing and consuming processes and in case of a process all input and output

flows. For example if the user clicks on the OILCRD commodity, the RES in Figure 2-43 is then

shown.

In this manner the user can cascade through the RES to better visualize the interrelationships and

competing process throughout the network.

Figure 2-42 RES view in VEDA-FE


47

Figure 2-43 RES view OILCRD

Note that at the bottom of the RES the data associated with the focus commodity or process can also

be displayed by means of the Show Data Cube button below the menu bar.

2.4 Generating and solving the TIMES model via VEDA-FE

2.4.1 Running and solving the model

In order to actually generate and solve the model, the data managed by VEDA-FE must be extracted

and prepared for the TIMES code31. To do so, from the FE navigator, the user submits a job run by

means of the “Solve” facility, initiated from the „Basic Functions‟ menu or by pressing the F9

function key. This then presents the Case Manager form, as shown in Figure 2-44.

Figure 2-44 VEDA-FE Case Manager for submitting model runs

31

In order for VEDA-FE to “talk” with GAMS it has to already be properly installed with the appropriate Windows

System PATH set, as described in the VEDA installation at www.kanors.com/vedasupport.



48

Using the Case Manager the user selects the scenarios (Base, B-NewTechs and SysSettings), the

region(s) (REG1), the end year (2001) and the model variant; with the name of the corresponding

run entered on the top line of the form.

In this Times Tutorial, the user will find in the paragraph 2.5 the analysis of three different runs:

- Times_Tutorial_Base (selects from the case manage the scenarios Base, B-NewTechs and

SysSettings);

- Times_Tutorial_IMP_Bound (selects from the case manage the scenarios Base, B-

NewTechs, IMP_Bound and SysSettings); for more details about the Scenario_IMP_Bound

see the paragraph 2.2.6.1;

- Times_Tutorial_CO2N_Bound (selects from the case manage the scenarios Base, B-

NewTechs, CO2N_Bound and SysSettings); for more details about the

Scenario_IMP_Bound see the paragraph 2.2.6.1;

The 1st time a user runs a model, the appropriate solver (e.g., CPLEX or XPRESS or MINOS)

should be selected via the pull-down list (showing CPLEX initially), and then it is advisable to click

on the solver name to set the options to be used. When the option form is opened (shown in Figure

2-45), the user selects the desired options (or simply leaves the default setting), and saves.

Figure 2-45 Cplex options

Other options on the Case Manager form allow the user to update the „RUN file‟, the GAMS

command file that launches the batch solution program (see Figure 2-46). For example the user

wants to see in the run output file (<case>.LST) the list of equations generated by this specific

TIMES model (LIMROW=xxxx), or turn on the solution listing (SOLPRINT=ON) 32

. In addition,

the user can control whether to close the Command window in which the model run takes place

when the run finishes, or leave it open.

32

More solve options are explained in the GAMS manual, downloadable from http://www.gams.com.

http://www.gams.com/


49

Once the appropriate selections have been made, simply click the „Solve‟ button in the Case

Manager form. The run will then be initiated. After the data is extracted from the database a

Windows Command Prompt window opens automatically and displays the progress of the matrix

generation, the solution and report writing of the TIMES model run, as shown in Figure 2-47.

Figure 2-46: Default value for LIMROW and LIMCOL

Figure 2-47: Command window showing the beginning and the optimal solution

To run the Times Tutorial and obtain the same results explained in the next paragraphs a user

should be selected the Model Variants from the FE Case Manager and in the window 'OBJ Function

Variant' set the options MODIFIED33

and then ok (shown in Figure 2-48).

33

See the ETSAP web site at http://www.etsap.org/Docs/TIMES-Objective-Variants.pdf for more information about

OBJ Function Variant

http://www.etsap.org/Docs/TIMES-Objective-Variants.pdf


50

Figure 2-48: Set Model Variants

2.4.2 Generation and solution files

2.4.2.1 Contents of the run folders

Model runs are preformed in the designated Wrk folder shown on the Case Manager form, normally

C:\VEDA\VEDA_FE\GAMS_WrkTIMES, but the user can designated different run folders for

different models if desired. When the run is finished34

, several new files appear in relevant

directories. These fall into four main groups:

- <scenario>.DD (Data Dictionary) files, created by VEDA-FE, assemble model inputs in text

files in GAMS syntax readied for the TIMES code;

- <run>.LST files, created by GAMS, echo the compilation and run information, as well as

report the model solution and equation listing, if requested, as well as report any errors or

infeasibilities;

- *.LOG files report information on the consistency of input data with respect to what TIMES

expects and/or requires and the solution run itself (logs produced by TIMES are stored in the

C:\VEDA\VEDA_FE\GAMS_WrkTIMES folder, those from VEDA-FE go to

C:\VEDA\VEDA_FE\LOGS), and

- <run>.VD* solution dump files prepare the model results for analysis with VEDA-BE.

The content and function of each file is explained hereafter for the simplified energy supply model

in example.

34

Here it is assumed the correct termination of the run, signaled, for instance, by the message „optimal solution found‟

in the DOS window (see above, Figure 2-47).


51

2.4.2.2 Data Dictionary files

As mentioned in the previous section, VEDA-FE extracts input data in a format compatible with the

TIMES GAMS code. A Data Dictionary (or DD) file is generated for each scenario involved in a

run, and stored in the run folder C:\VEDA\VEDA_FE\GAMS_WrkTimes. In the TIMES Tutorial

example, three DD files are generated (see Table 2.2) as part of the BASE run:

- BASE.DD containing the base year structure of the model (from the calibration templates)

and the demand projections;

- SysSettings.DD containing the inter/extrapolation settings and data adjustments, and

- SubRes.DD file containing the structure and data associated with the new technologies. For

the TIMES Tutorial example there are no new technologies, thus in this file there are only

few header rows.

If other SubRES or scenarios are involved in a run, VEDA-FE generates the associated DD files as

well.

The user rarely has a need or interest in viewing the DD files. The GAMS sets and parameters used

in the DD files are described in detail in the second part of the documentation of the TIMES model

(Reference or link to www.etsap.org/Docs/TIMESDoc-Details.pdf).

Table 2.2 Data Dictionary files for the TIMES Tutorial example (to be read by column)

BASE.dd

$ONEMPTY SET TOP

$ONEPS /

$ONWARNING REG1.EXPCO2N.CO2N.IN

$SET RUN_NAME '' REG1.IMPOILCRD1.OILCRD.OUT

$SET SCENARIO_NAME 'BASE' REG1.MINOILCRD1.CO2N.OUT

SET ALL_REG REG1.MINOILCRD1.OILCRD.OUT

/ REG1.MINOILCRD2.CO2N.OUT

IMPEXP REG1.MINOILCRD2.OILCRD.OUT

MINRNW REG1.TECTPES1.OILCRD.IN

REG1 REG1.TECTPES1.TPES.OUT

/ REG1.TECTPES2.OILCRD.IN

REG1.TECTPES2.TPES.OUT

SET REG /

/

REG1 'REG1' SET TOP_IRE

/ /

MINRNW.CO2N.REG1.CO2N.MINOILCRD1

SET CUR MINRNW.OILCRD.REG1.OILCRD.MINOILCRD1

/ MINRNW.CO2N.REG1.CO2N.MINOILCRD2

CUR MINRNW.OILCRD.REG1.OILCRD.MINOILCRD2

/ IMPEXP.OILCRD.REG1.OILCRD.IMPOILCRD1

REG1.CO2N.IMPEXP.CO2N.EXPCO2N

SET DATAYEAR IMPEXP.OILCRD.REG1.OILCRD.IMPNRGZ

/ IMPEXP.TPES.REG1.TPES.IMPDEMZ

2000 '2000' /

2001 '2001'

http://www.etsap.org/Docs/TIMESDoc-Details.pdf


52

/ SET PRC_DESC

/

SET MODLYEAR REG1.EXPCO2N 'Emissions Permits'

/ REG1.IMPDEMZ 'Dummy Import of DEM'

2000 '2000' REG1.IMPMATZ 'Dummy Import of MAT'

2001 '2001' REG1.IMPNRGZ 'Dummy Import of NRG'

/ REG1.IMPOILCRD1 'Import of Crude Oil'

REG1.MINOILCRD1 'Crude Oil domestic extraction'

SET TS_GROUP REG1.MINOILCRD2 'Crude Oil potential extraction'

/ REG1.TECTPES1 'Base year Demand technology'

REG1.ANNUAL.ANNUAL REG1.TECTPES2 'New Demand technology'

/ /

SET UNITS SET COM_DESC

/ /

CUR REG1.CO2N 'CO2 emission'

kt REG1.OILCRD 'Crude Oil'

Mt REG1.TPES 'Total primary energy supply'

PJ /

PJa

/ SET PRC_NOFF

/

SET UNITS_COM REG1.TECTPES2.BOH.2000

/ /

kt

PJ PARAMETER

/ ACT_BND ' '/

REG1.2000.MINOILCRD1.ANNUAL.UP 10000

SET UNITS_CAP REG1.2001.MINOILCRD1.ANNUAL.UP 9000

/ REG1.2000.MINOILCRD2.ANNUAL.UP 0

PJa REG1.2001.MINOILCRD2.ANNUAL.UP 3000

/ /

SET UNITS_ACT PARAMETER

/ REG1.2000.IMPNRGZ.CUR 9999

kt REG1.2000.IMPMATZ.CUR 9999

Mt REG1.2000.IMPDEMZ.CUR 999999

PJ /

/

PARAMETER

SET UNITS_MONY ACT_EFF ' '/

/ REG1.2000.TECTPES1.ACT.ANNUAL 1

CUR REG1.2000.TECTPES2.ACT.ANNUAL 1.3

/ /

SET COM_GRP PARAMETER

/ B ' '/

CO2N 2000 2000

OILCRD 2001 2001

TPES /

/

PARAMETER

SET COM COM_PROJ ' '/

/ REG1.2000.TPES 15000

CO2N REG1.2001.TPES 16000

OILCRD /


53

TPES

/ PARAMETER

D ' '/

SET COM_TMAP 2000 1

/ 2001 2

REG1.DEM.TPES /

REG1.ENV.CO2N

REG1.NRG.OILCRD PARAMETER

/ E ' '/

2000 2000

SET COM_UNIT 2001 2002

/ /

REG1.CO2N.kt

REG1.OILCRD.PJ PARAMETER

REG1.TPES.PJ FLO_EMIS ' '/

/ REG1.2000.MINOILCRD1.ACT.CO2N.ANNUAL 10

REG1.2000.MINOILCRD2.ACT.CO2N.ANNUAL 10

SET PRC /

/

EXPCO2N PARAMETER

IMPDEMZ IRE_PRICE ' '/

IMPMATZ REG1.2000.MINOILCRD1.OILCRD.ANNUAL.REG1.IMP.CUR 1

IMPNRGZ REG1.2001.MINOILCRD1.OILCRD.ANNUAL.REG1.IMP.CUR 1.1

IMPOILCRD1 REG1.2001.MINOILCRD2.OILCRD.ANNUAL.REG1.IMP.CUR 6

MINOILCRD1 REG1.2000.EXPCO2N.CO2N.ANNUAL.REG1.EXP.CUR -0.005

MINOILCRD2 REG1.2001.EXPCO2N.CO2N.ANNUAL.REG1.EXP.CUR -0.006

TECTPES1 REG1.2000.IMPOILCRD1.OILCRD.ANNUAL.REG1.IMP.CUR 4.9

TECTPES2 REG1.2001.IMPOILCRD1.OILCRD.ANNUAL.REG1.IMP.CUR 4.04

/ /

SET PRC_MAP PARAMETER

/ M ' '/

REG1.DMD.TECTPES1 2000 2000

REG1.DMD.TECTPES2 2001 2001

REG1.IRE.EXPCO2N /

REG1.IRE.IMPDEMZ

REG1.IRE.IMPMATZ PARAMETER

REG1.IRE.IMPNRGZ NCAP_COST ' '/

REG1.IRE.IMPOILCRD1 REG1.2000.TECTPES1.CUR 10

REG1.IRE.MINOILCRD1 REG1.2000.TECTPES2.CUR 12.75

REG1.IRE.MINOILCRD2 /

/

PARAMETER

SET PRC_TSL NCAP_TLIFE ' '/

/ REG1.2000.TECTPES1 3

REG1.EXPCO2N.ANNUAL REG1.2000.TECTPES2 3

REG1.IMPDEMZ.ANNUAL /

REG1.IMPMATZ.ANNUAL

REG1.IMPNRGZ.DAYNITE PARAMETER

REG1.IMPOILCRD1.ANNUAL VDA_EMCB ' '/

REG1.MINOILCRD1.ANNUAL REG1.2000.OILCRD.CO2N 70

REG1.MINOILCRD2.ANNUAL /

REG1.TECTPES1.ANNUAL

REG1.TECTPES2.ANNUAL

/


54

SET PRC_ACTUNT

/

REG1.EXPCO2N.CO2N.kt

REG1.IMPDEMZ.DEM.PJ

REG1.IMPMATZ.MAT.Mt

REG1.IMPNRGZ.NRG.PJ

REG1.IMPOILCRD1.OILCRD.PJ

REG1.MINOILCRD1.OILCRD.PJ

REG1.MINOILCRD2.OILCRD.PJ

REG1.TECTPES1.TPES.PJ

REG1.TECTPES2.TPES.PJ

/

B-NewTechs.dd

$ONEMPTY

$ONEPS

$ONWARNING

$SET RUN_NAME 'Times_Tutorial_Base'

$SET SCENARIO_NAME 'B-NewTechs'

SET TOP_IRE

/

/

SysSettings.dd

$ONEMPTY

$ONEPS

$ONWARNING

$SET RUN_NAME 'Times_Tutorial_Base'

$SET SCENARIO_NAME 'SysSettings'

SET TOP_IRE

/

/

PARAMETER

G_DRATE ' '/

REG1.2000.CUR 0.07

/

PARAMETER

G_YRFR ' '/

REG1.ANNUAL 1

/

2.4.2.3 LST file

The <run name>.LST file is produced by GAMS. It shows all the information about the TIMES

program compilation, the model generation, the matrix statistics, the solution, and run statistics.

Most importantly, it is the only place where any runtime errors or infeasibilities are reported one by

one35

, if encountered. The content and length of each section in the list file is different according to

the options indicated in the run file (explained above in paragraph 2.4.1). The LST file may be

35

And if the appropriate solver options are indicated. For details see the User‟ Guide of GAMS.


55

opened and viewed directly from the VEDA-FE Case Manager form by clicking the LST Files

button on the top of the form and selecting the desired LST file. The tables that follow show three

important sections of the LST file for the TIMES Tutorial model example.

Table 2.3 shows the complete list of model equations. LIMROW specifies the number of specific

equations per instance that are displayed in the listing file. Setting LIMROW to a high number

ensures that all primal equations are shown. Similarly, LIMCOL controls the number variables

which are reported per instance (VAR_ACT, VAR_FLO, etc.). Table 2.4 reproduces the solution

summary report. Table 2.5 list the solution of all primary and dual variables of the TIMES Tutorial

example model. The LEVEL values contain the primal solution, with the associated dual value

reported in the MARGINAL column. The user should be aware that the MARGINAL values

reported in the LST file are discounted values describing the associated cost changes for the entire

modelling horizon, while in VEDA-BE the undiscounted, annual dual variables are reported, other

than for the total discounted system cost. For example the marginal value 4.9 of the constraint

EQG_COMBAL REG1.2000.OILCRD.ANNUAL expresses that the discounted system cost would

increase by 4.9 Mio.$, if the oil production or demand would be increase or reduced by one unit in

the year 2000. If the period 2000 would comprise more than one year, the marginal value has to be

divided by the period duration and divided by the discounting factor to get the annual undiscounted

marginal values.

Table 2.3 Equation Listing

PRIMAL EQUATIONS (see TIMES Users’ Guide, part 2, chapter 5)36

---- EQ_OBJ =E= Overall Objective Function

EQ_OBJ.. VAR_OBJINV(REG1,CUR) + VAR_OBJFIX(REG1,CUR) + VAR_OBJVAR(REG1,CUR) -

VAR_OBJSAL(REG1,CUR) - OBJz =E= 0 ; (LHS = 0)

---- EQ_OBJFIX =E=

EQ_OBJFIX(REG1,CUR).. - VAR_OBJFIX(REG1,CUR) =E= 0 ; (LHS = 0)

---- EQ_OBJINV =E=

EQ_OBJINV(REG1,CUR)..10*VAR_NCAP(REG1,2000,TECTPES1)+9.34579439252336*VAR_NCAP(REG1,20

01,TECTPES1) + 11.9158878504673*VAR_NCAP(REG1,2001,TECTPES2) - VAR_OBJINV(REG1,CUR) =E= 0

; (LHS = 0)

---- EQ_OBJSALV =E=

EQ_OBJSALV(REG1,CUR).. 2.9070249129128*VAR_NCAP(REG1,2001,TECTPES1) +

3.70645676396382*VAR_NCAP(REG1,2001,TECTPES2) - VAR_OBJSAL(REG1,CUR) =E= 0 ; (LHS = 0)

---- EQ_OBJVAR =E=

EQ_OBJVAR(REG1,CUR)...0.005*VAR_ACT(REG1,2000,2000,EXPCO2N,ANNUAL)+8888*VAR_ACT(REG1,

2000,2000,IMPDEMZ,ANNUAL)+4.9*VAR_ACT(REG1,2000,2000,IMPOILCRD1,ANNUAL)+VAR_ACT(REG

1,2000,2000,MINOILCRD1,ANNUAL)+0.0108481090051533*VAR_ACT(REG1,2001,2001,EXPCO2N,ANNUA

L)+16069.6654729671*VAR_ACT(REG1,2001,2001,IMPDEMZ,ANNUAL)+7.30439339680321*VAR_ACT(RE

G1,2001,2001,IMPOILCRD1,ANNUAL)+1.9888199842781*VAR_ACT(REG1,2001,2001,MINOILCRD1,ANNU

AL)+10.8481090051533*VAR_ACT(REG1,2001,2001,MINOILCRD2,ANNUAL) - VAR_OBJVAR(REG1,CUR)

=E= 0 ; (LHS = 0)

---- EQL_CAPACT =L= Capacity Utilzation (=L=)

36

In the same chapter you find the (complex) formula for discounting and salvaging.


56

EQL_CAPACT(REG1,2000,2000,TECTPES1,ANNUAL)..VAR_ACT(REG1,2000,2000,TECTPES1,ANNUAL) -

VAR_NCAP(REG1,2000,TECTPES1) =L= 0; (LHS = 0)


VAR_NCAP(REG1,2000,TECTPES1)-VAR_NCAP(REG1,2001,TECTPES1) =L= 0 ; (LHS = 0)


VAR_NCAP(REG1,2001,TECTPES2) =L= 0 ; (LHS = 0)

---- EQG_COMBAL =G= Commodity Balance (=G=)

EQG_COMBAL(REG1,2000,CO2N,ANNUAL).. -

VAR_ACT(REG1,2000,2000,EXPCO2N,ANNUAL)+10*VAR_ACT(REG1,2000,2000,MINOILCRD1,ANNUAL)

+70*VAR_ACT(REG1,2000,2000,TECTPES1,ANNUAL)+VAR_IRE(REG1,2000,2000,MINOILCRD1,CO2N,AN

NUAL,IMP) =G= 0 ; (LHS = 0)

EQG_COMBAL(REG1,2000,OILCRD,ANNUAL)..

VAR_ACT(REG1,2000,2000,IMPOILCRD1,ANNUAL)+VAR_ACT(REG1,2000,2000,MINOILCRD1,ANNUAL)

-VAR_ACT(REG1,2000,2000,TECTPES1,ANNUAL) =G= 0 ; (LHS = 0)

EQG_COMBAL(REG1,2000,TPES,ANNUAL)..

VAR_ACT(REG1,2000,2000,IMPDEMZ,ANNUAL)+VAR_ACT(REG1,2000,2000,TECTPES1,ANNUAL) =G=

15000 ; (LHS = 0, INFES = 15000 ****)

EQG_COMBAL =G= Commodity Balance (=G=)

EQG_COMBAL(REG1,2001,CO2N,ANNUAL).. -

VAR_ACT(REG1,2001,2001,EXPCO2N,ANNUAL)+10*VAR_ACT(REG1,2001,2001,MINOILCRD1,ANNUAL+

10*VAR_ACT(REG1,2001,2001,MINOILCRD2,ANNUAL)+70*VAR_ACT(REG1,2001,2001,TECTPES1,ANNU

AL)+53.8461538461538*VAR_ACT(REG1,2001,2001,TECTPES2,ANNUAL)+VAR_IRE(REG1,2001,2001,MIN

OILCRD1,CO2N,ANNUAL,IMP)+VAR_IRE(REG1,2001,2001,MINOILCRD2,CO2N,ANNUAL,IMP) =G= 0 ;

(LHS = 0)

EQG_COMBAL(REG1,2001,OILCRD,ANNUAL)..

VAR_ACT(REG1,2001,2001,IMPOILCRD1,ANNUAL)+VAR_ACT(REG1,2001,2001,MINOILCRD1,ANNUAL+

VAR_ACT(REG1,2001,2001,MINOILCRD2,ANNUAL)-VAR_ACT(REG1,2001,2001,TECTPES1,ANNUAL)-

0.769230769230769*VAR_ACT(REG1,2001,2001,TECTPES2,ANNUAL) =G= 0 ; (LHS = 0)

EQG_COMBAL(REG1,2001,TPES,ANNUAL)..

VAR_ACT(REG1,2001,2001,IMPDEMZ,ANNUAL)+VAR_ACT(REG1,2001,2001,TECTPES1,ANNUAL)+VAR

_ACT(REG1,2001,2001,TECTPES2,ANNUAL) =G= 16000 ; (LHS = 0, INFES = 16000 ****)

DUAL EQUATIONS (price formation equations) by Variable

VAR_ACT(REG1,2000,2000,EXPCO2N,ANNUAL)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)

0.005 EQ_OBJVAR(REG1,CUR)

-1 EQG_COMBAL(REG1,2000,CO2N,ANNUAL)

VAR_ACT(REG1,2000,2000,IMPDEMZ,ANNUAL)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)

8888 EQ_OBJVAR(REG1,CUR)

1 EQG_COMBAL(REG1,2000,TPES,ANNUAL)

VAR_ACT(REG1,2000,2000,IMPOILCRD1,ANNUAL)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)


1 EQG_COMBAL(REG1,2000,OILCRD,ANNUAL)

VAR_ACT(REG1,2000,2000,MINOILCRD1,ANNUAL)

(.LO, .L, .UP, .M = 0, 0, 10000, 0)

1 EQ_OBJVAR(REG1,CUR)

10 EQG_COMBAL(REG1,2000,CO2N,ANNUAL)


VAR_ACT(REG1,2000,2000,TECTPES1,ANNUAL)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)

1 EQL_CAPACT(REG1,2000,2000,TECTPES1,ANNUAL)



57

-1 EQG_COMBAL(REG1,2000,OILCRD,ANNUAL)


VAR_ACT(REG1,2001,2001,EXPCO2N,ANNUAL)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)


-1 EQG_COMBAL(REG1,2001,CO2N,ANNUAL)

VAR_ACT(REG1,2001,2001,IMPDEMZ,ANNUAL)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)



VAR_ACT(REG1,2001,2001,IMPOILCRD1,ANNUAL)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)




(.LO, .L, .UP, .M = 0, 0, 9000, 0)





(.LO, .L, .UP, .M = 0, 0, 3000, 0)





(.LO, .L, .UP, .M = 0, 0, +INF, 0)



-1 EQG_COMBAL(REG1,2001,OILCRD,ANNUAL)



(.LO, .L, .UP, .M = 0, 0, +INF, 0)


53.8462 EQG_COMBAL(REG1,2001,CO2N,ANNUAL)

-0.7692 EQG_COMBAL(REG1,2001,OILCRD,ANNUAL)


VAR_IRE(REG1,2000,2000,MINOILCRD1,CO2N,ANNUAL,IMP)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)



(.LO, .L, .UP, .M = 0, 0, +INF, 0)



(.LO, .L, .UP, .M = 0, 0, +INF, 0)


VAR_NCAP(REG1,2000,TECTPES1)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)

10 EQ_OBJINV(REG1,CUR)

-1 EQL_CAPACT(REG1,2000,2000,TECTPES1,ANNUAL)



(.LO, .L, .UP, .M = 0, 0, +INF, 0)

9.3458 EQ_OBJINV(REG1,CUR)

2.097 EQ_OBJSALV(REG1,CUR)



(.LO, .L, .UP, .M = 0, 0, +INF, 0)

11.9159 EQ_OBJINV(REG1,CUR)

3.7065 EQ_OBJSALV(REG1,CUR)


VAR_OBJINV(REG1,CUR)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)


58

1 EQ_OBJ

-1 EQ_OBJINV(REG1,CUR)

VAR_OBJFIX(REG1,CUR)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)

1 EQ_OBJ

-1 EQ_OBJFIX(REG1,CUR)

VAR_OBJVAR(REG1,CUR)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)

1 EQ_OBJ

-1 EQ_OBJVAR(REG1,CUR)

VAR_OBJSAL(REG1,CUR)

(.LO, .L, .UP, .M = 0, 0, +INF, 0)

-1 EQ_OBJ

-1 EQ_OBJSALV(REG1,CUR)

OBJz

(.LO, .L, .UP, .M = 0, 0, +INF, 0)

-1 EQ_OBJ

Table 2.4 Solution Report

Solve Summary

MODEL TIMES OBJECTIVE OBJz

TYPE LP DIRECTION MINIMIZE

SOLVER CPLEX FROM LINE 46152

SOLVER STATUS 1 NORMAL COMPLETION

MODEL STATUS 1 OPTIMAL

OBJECTIVE VALUE 259968.903116

Table 2.5 Solution list

EQUATIONS LOWER LEVEL UPPER MARGINAL

---- EQ_OBJ Overall Objective Function

EQU EQ_OBJ - - - -1

EQU EQ_OBJFIX REG1.CUR - - - -

EQU EQ_OBJINV REG1.CUR - - - -1

EQU EQ_OBJSALV REG1.CUR - - - 1

EQU EQ_OBJVAR REG1.CUR - - - -1

---- EQU EQL_CAPACT Capacity Utilzation (=L=)

REG1.2000.2000.TECTPES1.ANNUAL -INF - - -3.5612



---- EQU EQG_COMBAL Commodity Balance (=G=)

REG1.2000.CO2N .ANNUAL - 1150000 +INF -

REG1.2000.OILCRD.ANNUAL - - +INF 4.9000

REG1.2000.TPES .ANNUAL 15000 15000 +INF 8.4612

REG1.2001.CO2N .ANNUAL - 1210000 +INF -

REG1.2001.OILCRD.ANNUAL - - +INF 7.3044

REG1.2001.TPES .ANNUAL 16000 16000 +INF 13.7432

VARIABLES

---- VAR VAR_ACT

REG1.2000.2000.EXPCO2N .ANNUAL +INF 0.0050

REG1.2000.2000.IMPDEMZ .ANNUAL - - +INF 8879.5388

REG1.2000.2000.IMPOILCRD1.ANNUAL - 5000 +INF -

REG1.2000.2000.MINOILCRD1.ANNUAL - 10000 10000 -3.9000

REG1.2000.2000.TECTPES1 .ANNUAL - 15000 +INF .


59

REG1.2001.2001.EXPCO2N .ANNUAL +INF 0.0108

REG1.2001.2001.IMPDEMZ .ANNUAL - - +INF 16005.9223

REG1.2001.2001.IMPNRGZ .ANNUAL - - +INF 18071.0693

REG1.2001.2001.IMPOILCRD1.ANNUAL - 7000 +INF -

REG1.2001.2001.MINOILCRD1.ANNUAL - 9000 9000 -5.3156

REG1.2001.2001.MINOILCRD2.ANNUAL - - 3000 3.5437

REG1.2001.2001.TECTPES1 .ANNUAL - 16000 +INF -

---- VAR VAR_IRE

REG1.2000.2000.MINOILCRD1.CO2N.ANNUAL.IMP - - +INF EPS



---- VAR VAR_NCAP

REG1.2000.TECTPES1 - 15000 +INF -

REG1.2001.TECTPES1 - 1000 +INF -

REG1.2001.TECTPES2 - - +INF -

---- VAR VAR_OBJINV REG1.CUR - 159345.7944 +INF -

---- VAR VAR_OBJFIX REG1.CUR - +INF 1.0000 -

---- VAR VAR_OBJVAR REG1.CUR - 103530.1336 +INF -

---- VAR VAR_OBJSAL REG1.CUR - 2907.0249 +INF -

---- VAR OBJz -INF 259968.9031 +INF -

2.4.2.4 Log files

The QA_Check.LOG file is a quality assurance log file where checks related to the correctness of

your model specification with respect to what TIMES expects/requires are reported. In this file

errors and warnings are reported with a description of the error/warning severity. This file is stored

in C:\VEDA\VEDA_FE\GamsWrkTimes. When 1st assembling a model the user should review the

file to ensure the basic correctness of the model as presented to TIMES.

VEDA-FE also produces log files during the process of generating the Data Dictionary files

associated with model runs, as well as the import activities managed by VEDA-Navigator. These

log files, as well as error files that may be generated by VEDA-FE itself if internal system errors

occur, are stored in the folder C:\VEDA\VEDA_FE\Logs (if you have installed the software in the

default VEDA system folder), see Figure 2-49. The Log files may be viewed directly from the

VEDA-FE Case Manager form from the list presented by means of the Log Files button on the top

of the form.

Figure 2-49 Files in the VEDA-FE Log folder

These files, generated by VEDA-FE, give information about errors while import templates,

commodities declaration, missing commodities, problems with process commodities group


60

definition, problems related to the scenario files, inconsistencies in the model structure, etc.. Some

of these errors should be solved by the user (e.g. missing commodities), but some others are related

to the software (e.g. ErrorReportVEDA.doc). In this latter case the ...\Logs folder files should be

sent to the VEDA system administrator to assist with diagnosing the nature of any problem

encountered with the software itself. They are not shown here because for the TIMES Tutorial

model they are empty.

2.4.2.5 VEDA-BE solution files

Upon the successful completion of a model run, the following <run>.VD* files are created by the

TIMES GAMS code that will be feed to VEDA-BE to review the results of the model run. These

consist of:

- VD – data file with the actual run results, both primal and dual;

- VDS – the user Sets;

- VDE – the elements of each of the Sets, and

- VDT – the topology information for the RES (actually created by VEDA-FE during

generation of the Data Dictionaries at run time).

These files are stored in the folder C:\VEDA\VEDA_FE\GAMSWrkTimes. They contain all the

model results, and associated information (e.g., sets, topology) to be passed to VEDA-BE for

analysis by the user. A thorough description of VEDA-BE will be given below.

2.5 Analyzing the model results with VEDA-BE

The second37

component of the VErsatile Data Analyst, called Back End (VEDA-BE) is used for

the analysis of model results, as well as being able to accept VEDA-FE input if desired. VEDA-BE

relies on sets (both TIMES standard sets as well as user-defined (re-)grouping sets defined in

VEDA-BE), and user-defined tables which present the data as dynamic “cubes” or pivot tables. This

section illustrates how to import the results, select pre-defined tables, build new tables of interest,

define new user sets and take advantage of the pivotal functionality of the resulting cube tables.

2.5.1 Results import in VEDA-BE

In order to analyse the results of a model run you need to import the scenario run solution files into

VEDA-BE. To do so the user needs to follow these steps:

1) Launch VEDA-BE by clicking the corresponding “red book” icon;

2) From the starting window, select „File‟ and „Open Database‟, as shown in Figure 2-50;

37

In fact VEDA-BE was been designed and completed long before VEDA-FE. VEDA-BE is also used with ANSWER,

and in fact other GAMS models (e.g., GEMe3).


61

3) In the new window select the path C:\VEDA\VEDA-BE\Databases\TUTORIAL, as shown

in Figure 2-51, check that the lower right line contains the desired subfolder name, and

4) Import the run just completed (e.g. Times_Tutorial_Base) by selecting „Results‟,

„Import/Archive‟ or just press F7 (see Figure 2-53), check the run just made (e.g.

Times_Tutorial_Base) and hit the OK button. Note that runs previously loaded for which

newer results are found will be pre-selected by VEDA-BE, as it is assumed they should be

re-imported.

Figure 2-50 VEDA-BE results import

Figure 2-51 Path database TIMES_TUTORIAL

Figure 2-52 VEDA-BE Import/Archive window

When the results are correctly loaded, the main VEDA-BE selection form opens as presented in

Figure 2.28. The screen is divided into two parts: the table selection list on the left, and the table

specification component section on the right. The selection list allows the user to select previously

created table from the pull-down list or prepare to define a new table. The tabs on the specification

area correspond to the various aspects of a table, broken into the Attributes that correspond to the

various model results, and the indexes that qualify the information associated each component, as

well as the desired scenarios (all by default for all dimensions). Two tabs deserve special attention,


62

the Commodity and Process tab, in that they have two sections, the upper part listing the sets (both

model and user) and the lower the individual items. More explanation of the tabs will be provided

later.

Figure 2-53 VEDA-BE main table specification form

2.5.2 Results tables

Once the results have been loaded into VEDA-BE a series of predefined tables, discussed in the

next sections, can be used to examine the results.

After requesting a table the user can rearrange the layout of the cube by adding/removing

dimensions (columns and rows) to the table by dragging and dropping components from/to the area

above the current row designator columns. When the cursor is placed over a dimension a crosshair

appears. Then, holding the left mouse button down and sliding to a new position a green line will

appear indicating that the dimension may now be “dropped” there. Any dimension not positioned as

part of the row/column table layout definition appears at the top of the page. These dimensions have

their values summed in the cube. For each dimension on top of the page if more than 1 value exists

for that dimension its name will be displayed between two asterisks. [Note that for any dimension

where only a single value exists said dimension is automatically moved up top.]

Upon exiting the cube view of the user will be prompted to resave the table since its layout has been

changed. The last saved layout will be that used the next time the table is displayed.

2.5.2.1 Check dummy imports

The first table to examine is „Check dummy imports.‟ This table should be empty in a healthy

model, as it reports any „dummy‟ (backstop) options that had to come into the solution in order to

avoid infeasibilities. Select the table from the „Table Definition‟ pull-down list and then click the

“View Table(s)” button to see the resulting data cube. If anything is shown in this table it means


63

that the model could not meet all the constraints imposed by the run, and would have been

infeasible without the dummies. This table can help the user to understand where there is a problem

in the model, assisting with the sometimes challenging task of sorting out the nature of the

inconsistency.

2.5.2.2 All prices

In this table (see Figure 2-54) are shown the shadow prices of fuels and demands. This table

reorders and undiscounts38

the values shown in the column MARGINALS of the section

EQUATIONS of the solution list (Table 2.5).

From the mathematical point of view the shadow price of a constraint is the incremental value of

the objective function per unit increase of that constraint‟s right hand side (RHS). From the

economic point of view, the shadow price is the marginal value of a resource (where each constraint

defines a scarcity “resource”). For instance, the marginal price of a commodity balance constraint is

not necessarily the marginal cost of producing that commodity. Indeed, when the RHS of the

balanced constraint is increased by one unit, one of two things may occur: either the system

produces one more unit of commodity, or else the system consumes one unit less of commodity

(perhaps by choosing more efficient end-use devices or by reducing a commodity-intensive energy

service, etc.). In other words, the marginal price of a commodity can be determined by technologies

on the production or the consumption side of the commodity.

In this Tutorial problem the prices of OILCRD in the year 2000 is always equal to cost of import,

which is the marginal option, in the period 2001-2 it is equal to the cost of import unless the energy

independence is pursued and the model is forced to use the more expensive domestic source. In this

simple case the marginal price of the energy demanded by the users (TPES) is equal to the price of

OILCRD plus the unit cost of the marginal user process (see next paragraph). This price collects all

the costs in the chain and reflects in a synthetic way the cost of different policies. CO2 is priced

only in the CO2 constrained case; the marginal price is equal to the cost of purchasing a permit; the

negative sign indicates that increasing the amount of CO2 by one unit reduces the total system cost.

In TIMES the shadow prices of commodities play a very important diagnostic role. If some shadow

price is clearly out of line (i.e. if it seems much too small or too large compared to the anticipated

market prices shown in Table 2.1), this indicates that the model‟s database may contain some errors.

For instance, if the shadow price of a commodity is zero and the quantity supplied is non zero, as

pointed by the second theorem of Linear Programming, it means that there is more supply that

demand of that commodity. The examination of shadow prices is just as important as the analysis of

38

The undiscounted marginal prices in the year 2000 are equal to discounted values because all economic values in this

model are expressed in euro of January 1 of the year 2000. The values of the following period refer to the centre of the

two year period 2001-2, which is January 1 of the year 2002. Given a discount rate r of 7% in real terms, the discount

factor is d=1/(1+r) has the value 0.93458. In order to discount the prices of Jan 1, 2002 back to Jan 1, 2000, the prices

are multiplied by the factor 0.87344.


64

the quantities produced and consumed of each commodity and of the technological investments. In

this example the shadow prices are in line with the market prices.

Figure 2-54 Prices by commodity, scenario and year

2.5.2.3 Demands

In Figure 2-55 the demand (TPES) for each run/period is shown. The table also shows what part of

the demand is satisfied by technology TECTPES1 and TECTPES2 (Process label). The values are

taken from the second column (LEVEL) of the corresponding variables in Table 2.5.

Figure 2-55 Demands by process, year and scenario

Why in the base case the additional demand of TPES in 2001 is covered by process TECTPES1?

Because the cost of satisfying the demand with TECTPES1 is 7.60 €/PJ (price of input oil + annuity

= 4.04+10*0.35612) is lower than the cost of TECTPES2 ( = 4.04/1.3+12.75*0.356123)

2.5.2.4 Total supply

Figure 2-55 shows the total supply for each period with the individual import and mining details.

The table reports in ordered way the values of the column level of the corresponding variables in

Table 2.5. Here a total was added to the table, and can be removed, by placing the cursor in the

column to be summed, pressing the right-mouse, and requesting/unselecting Total.


65

Figure 2-56 Total supply

2.5.2.5 Creating a new table - Emissions

A table is fully defined in two steps by:

a) selecting the contents of the table, i.e., what results that are to be presented in the cube,

and assigning the units (if desired), and

b) arranging the layout for the table, i.e., organizing the table‟s rows and columns.

Creating a new table requires selecting which types of results to show (attribute), and perhaps which

subset of processes, commodities, regions, scenarios, time-periods, etc. are to be included. This is

accomplished by following these steps to make a table that reports the CO2 emissions:

1) Click the down arrow for Table Definition (top-left) and select “New Table” (or use the

shortcut Crtl+N, or hit the button below the table list);

2) Select the Attribute tab in the specification window on the right and select VAR_FOUT (see

Figure 2-57), which reflects the output from processes and VAR_FIN to include the

emissions associated with exports (or sequestration);

3) Select the Set ENV (see Figure 2-58) from the upper window on the Commodities tab so

that the table displays all of the emissions tracked in the model, alternatively from the

Commodity list CO2N which corresponds only to CO2 emissions;

4) Pull-down the Unit list and select kt;

5) Click View Table(s) and provide the table name (e.g. Emissions) then click OK.

6) The resulting Emissions table can then be arranged as described previously to show the CO2

emission (CO2N) in kt per for each period associated with each process from which

emissions emanate (see Figure 2-59).


66

Figure 2-57 Selecting the VAR_FOUT attribute

Figure 2-58 Selecting Commodity Set ENV

Figure 2-59 Emissions by source, year and scenario

Only the total can be found in the Solution list (Table 2.5, section EQUATIONS, EQU_COMBAL,

CO2N, column LEVEL) and the amount of emission permits purchased from abroad (same table,


67

section VARIABLES, EXPCO2N, column LEVEL). The other values are calculated by VEDA-BE,

multiplying the values of the variables reported in the solution list (in Table 2.5) by the emission

coefficients given in input (and reported here in Figure 2-59).

2.5.2.6 All costs

The next important aspect is the cost. The table „All costs‟ presents all the undiscounted system

costs for mining, import and technologies39

. By using the drag-and-drop function and the right

mouse the results are rearranged as desired. For example Figure 2-60, shows the „Scenario‟ (as the

first outermost row qualifier) and „Period‟ (as the second outermost column in the table). In this

way it is possible to see all undiscounted costs in each period.

The table „All costs‟ presents all the undiscounted system costs (as shown in Figure 2-60) for

mining, import and technologies.

In Figure 2-60, the Vintage and Region are removed from the table (aggregated if >1 element), with

the Scenario name as first, the Attribute as second, the Commodity as third and the Process as the

fourth row qualifiers of the table. For the column only the Period label is chosen.

Figure 2-60 All costs table by variable, scenario and time period

39

The original table is useful for a first check but it is not shown here because it aggregates over the years undiscounted

costs.


68

None of the values shown in this table appears directly in the solution list. The cost of mining,

import and export (attribute Cost_Flow in Figure 2-60) are the multiplication of the level of the

variables (given in Table 2.5 for the base case) by their input costs (reported above in Table 2.1 and

Figure 2-16 or in Figure 2-30). The cost of using energy to satisfy the demand is in this case due

only to the purchase of the processes (related to processes TECTPES1 and 2). This is an investment

(attribute Cost_Inv in Figure 2-60). The cost in the year 2000 is given by

Cost_Inv(TECTPES1)=Quantity*Invcost*CRF

Where: Quantity is the LEVEL of the variable TECTPES1 (in Table 2.5)

Invcost is the unit purchase price of TECTPES1 (see Figure 2-18)

CRF is the Capital Recovery Factor = (1/(1+r))*r/(1-(1+r)-L

)= 0.356123, where

r=0.07 is the discount rate (Figure 2-10) and L=3 is the life time of the process

TECTPES1 (values in Figure 2-18)

Salvage represents the value of the investments at the end of the time horizon discounted to the base

year. In this simple case it is (1000*10*0.356123*(0.93458)3).

2.5.2.7 Total System Cost

The final key result of interest is the total system cost of the model. Select the table „_SysCost‟

from the „Table Definition‟ pull-down list and then click the “View Table(s)” button to see the

resulting data cube, shown in Figure 2-61 below. The costs shown in this table under the attribute

“ObjZ” are the total system costs discounted with the general discount rate to the base year of the

model, here the year 2000, the same value reported in the last line of Table 2.5. In this case the

values shown in Figure 2-61 can be recalculated by hand.

The OBJ value in the base case is the sum of all yearly costs depreciated to the base year (2000)

with the discount factor α=1/(1+t) which in this example is 0.93458. The same value is given by the

sum of the 3 OBJ function values reported in Table 2.5 (OBJINV, OBJVAR and OBJSAL).

Figure 2-61 _SysCost Table


69

2.5.3 Evaluating the effect of possible choices

The results show that the system develops differently according to the objectives (Table 2.6). In the

strictest financial perspective without additional policies, importing oil from abroad is the cheapest

way to satisfy the demand of the system if the existing domestic oil is not enough. The system has

two options for reducing imports: the development of domestic resource and the use of more

efficient consumer devices. Furthermore, the system has two options for reducing emissions: the

purchase of emission permits from abroad and the use of more efficient consumer devices.

In this example the energy efficiency improvement option (TECTPES2) is "robust", i.e. it is

convenient for reducing both emissions and energy dependence. In fact, in both scenarios it is the

next best, because, it is cheaper than buying emission permits or importing oil. The extra cost of

using TECTPES2 instead of TECTPES1 in 2001 is 1029 Beuro; in the same year it saves 104838

bbl of oil which are equivalents to saving 0.932 Beuro of imported oil and 15 MtCO2 of emissions

which are worth 0.162 Beuro. The combined value of the savings (oil and emissions) is higher than

the cost.

But the dependence and emissions can be further reduced if the more expensive domestic resources

are made available. The extra cost could be justified by economic or strategic consideration not

included in the model. For instance, if domestic resources can be developed with domestic labour

and technologies, the economic multiplier could be high enough (> 1.48) to bring back to the

domestic economy indirect economic benefits greater than the outlays for buying foreign oil.

Furthermore, higher cost for developing now domestic oil resources could be a rational way to

ensure against the risks of higher international oil prices in the longer term.

Table 2.6 Some scenario indicators

Indicator Scenario Base Case Mitigation Oil indepence

(Least cost) Bounds on emis. Bounds on import

Costs in 2001 Beuro/a 95160 95470 99654

Emissions in 2001 MtCO2/a 1210.0 1193.8 1216.5

Oil import in 2001 Mbbl/d 3.180 3.075 2.044

Oil dependence % 43.8% 42.5% 28.3%

The purchase of emission permits at 10€/tCO2 can reduce official emissions and avoid paying the

fee for non compliance a mitigation targets.

Other policies can be explored with this simple model could try to evaluate how much you could be

spent to avoid decreasing the production of existing domestic oil resources or of separating and

reinjecting the CO2 come out of the third oil fields or putting in the market a more efficient option

of satisfying the demand. Another way to gain more in depth understanding of the methodology is

to explore the effects of changing some input parameter, i.e. the effects of uncertainties. Eventually

the model can help you calculate the cost (or efficiency) goal of making TECTPES2 competitive

also in the base case or the surplus of the old domestic oil producers.


70

2.5.4 Setting Defaults40

Prior to defining and using tables, it is recommended to set a few default preferences to suit your

installation and tastes41

. Setting default options starts with selecting Tools in the main menu. Here

are a few useful defaults to set. Note that all these defaults may be superseded at later stages of table

definition. On the Tools/Options/General tab, choose the number format and whether or not you

wish to save cube files. Saving cube files is recommended since it speeds up the presentation of

repeatedly selected analysis tables (when no information is changed in the table since the last

request). On Tools/Options/Export Options, in the Export Path window, select the folder where you

want to save the exported files. You may also create a new folder into which exported files will be

saved by means of the standard new folder button (section 2.5.5 describes the export function).

Cube files may be exported in various formats (Excel, Word, HTML, text).

Tools/Options/Dimensions, and you may choose whether elements in each dimension are shown

according to their (short) code names or by their (long) descriptions. You may also choose whether

or not subtotals are shown for each particular dimension.

Figure 2-62 VEDA-BE setting options

2.5.5 Exporting result tables

Data presented in the current View may be exported in various formats. Two methods are available.

First method:

- Click on one of the Export to Excel/Word/HTML/Text icons at the top of the window. This

operation will create a file of the requested type in a pre-specified directory (section 2.5.4).

- The file will be automatically named by appending the date and time at which the filed was

saved to the table‟s name.

40

For more details see the VEDA-BE manual available at http://www.kanors.com/userguidebe.htm. 41

Note that in all cases VEDA does provide defaults of all options.

http://www.kanors.com/userguidebe.htm


71

Upon completing the Export operation VEDA will prompt the user to inidcate if the exported

information is to be presented for review.

Second method

Another way to copy all or part of a table consists of right-clicking while the cursor lies in the

numerical area of the table, and choosing the Copy Data. The table is then put in the Windows

clipboard, and may then be pasted to an application of your choice. This manner of operating is

often more convenient when debugging a run‟s results or if the tables must be pasted at specific

locations in a workbook or document for further processing. Note that the name, headings, and units

of the table are also copied.

Both of these methods also allow exporting a portion of the table, by first blocking the rectangular

area to be exported, and then only right-clicking to choose the export option.

2.5.6 Graphs

A portion of a table may be conveniently graphed as described here:

- Block a rectangular area of interest in the table;

- Click on the Graph icon on top of the VIEW window; a preliminary (line) graph is

presented, which may now be modify by swapping axes, changing the chart type to bar, line,

pie, area, etc., as well as two or three dimensional charts;

- the graph may be copied to the clipboard for later printing or importing into a document, and

- Series may be removed from a graph, and the scale is automatically refreshed (which may

be useful if series with different orders of magnitude co-exist in a graph).


72

3 Next Step: the TIMES_Demo Model

Actual energy systems are not as simple as shown in the previous chapter. They normally include

all energy flows, from solid fossil to natural gas and oil products, from central electricity to

distributed heat, ultimately delivering energy services such as space heating/cooling, lighting,

process heat, passenger miles travelled. Furthermore they include all the chains of technologies

from extraction to processing, from distribution to end-users and those delivering the energy

services. For a useful policy planning model, although not all details can be represented, there must

be enough detail to gain appropriate insight into the problems, options and future implications of

decisions. The previous sections served to provide the foundation with respect to what is involved

to build such a model. In this chapter a step towards working with a still simplified but more

complete model that begins to cover the breath of the energy system will be undertaken. To do so

some the aspects of the TIMES_DEMO model (section 3.1) will be presented and used to illustrate

how to represent such a system as a TIMES model built with VEDA (section 3.2).

The complete set of templates associated TIMES_DEMO model are distributed with the VEDA

installation.

3.1 Main aspects of the TIMES_DEMO model

3.1.1 Installation of the model

Here the simple steps necessary to open, import and run the DEMO model are presented:

1) Launch VEDA-FE, select the path in which the TIMES_DEMO is stored (as shown in the

paragraph 2.3.2), and the VEDA-Navigator (see Figure 3.1) with the TIMES_DEMO

templates presented is presented;

2) If any of the templates or database symbols indicate there are inconsistencies (shade light

red) click SYNC;

3) Open the Case Manager form (Figure 3-1), by selecting “Solve” from the Basic Functions

menu or pressing (F9), and

4) Click <Solve>, opening a Command window as shown in Figure 3-2 where the model is

actually generated and solved.

As the Demo model is always evolving it might be the case that the results are different

from that shown at the row Objective in the Figure 3-2 below (49974145), but it should be

similar.


73

Figure 3-1 VEDA-FE Navigator and Case Manager forms of the TIMES_DEMO model

Figure 3-2 Command window

The results must then be imported into VEDA-BE to be examined, by42

:

1) Launching VEDA-BE;

2) Importing the results by means of the menu (Results --> Import/Archive) or (F7);

3) Using Manage Input File Location to point to the folder where the run was carried out, and

4) Checking table _Syscost to see that the value for the run are the same as the DEMO_ BASE

already loaded.

42

See also at www.kanors.com/vedasupport.



74

3.1.2 Directory structure43

The files of the TIMES_DEMO are organized under VEDA_Models\TIMES_DEMO in three sub-

folders. In the TIMES_DEMO folder there are five base year templates, the SysSetting and

BY_Trans files:

- SysSettings - declares the default interpolation rules and parameter values, as well as

Commodity group definition44

(rule-based).

- BY_Trans - (B-Y Transformation file) this file supports all tables and works just like a

scenario file, with one important difference: the process and commodity filters see only

those elements that come from the B-Y templates.

- Base Year Templates45

with the current energy system details:

VT_DEMOT_ELC_V5 - power sector

VT_DEMOT_IND_V5 - industry

VT_DEMOT_OTH_V5 - agriculture, commercial and residential

VT_DEMOT_SUP_V5 - supply and upstream

VT_DEMOT_TRA_V5 - transportation

In addition, three sub-folders containing six excel files (as shown in the Figure 3-3)46

complete the

picture:

1) In the folder „Databases‟ VEDA-FE builds and updates the internal ACCESS database of

the model; normally users are not concerned with these internal databases (and should not

access this folder unless);

2) The folder „SubRes_TMPL‟ contains rule-based data specification (file SubRES_B-

NewTechs.xls) and transformation (for the respective SubRES) for new technologies;

3) In the folder „SuppXLS‟ rule-based data specifications and transformations for the entire

Reference Energy System are specified by means of three sub-folders (Demands, Trades and

UConstraints) and three Excel files:

a) Demand - driver files with demand drivers (DEMO_Base), and sensitivity series and

region-segment driver allocation table (Dem_Alloc+Series).

b) Trades - trade scenario file (ScenTrade_TRADEAttribs) with parameters for the

inter-regional trade technologies; this is the only way to create parameters for the

inter-regional exchange processes (rule-based).

c) UConstraints - user constraints scenario file (ScenUC_Test_AFS_UC) with special

share constraints; big set based on output and the subset based on input (rule-based).

d) Excel files in SuppXLS

43

See also paragraph 2.2 44

For more details see paragraph 3.2.2. 45

Note that users can declare their model through as many Base Year templates as desired. 46

Each model is organized with a similar sub-folder structure


75

- Scen_CO2_Tax and Scen_UCTest - are two examples of scenario files

containing the introduction of the tax on CO2 emissions and a set of illustrative

user-defined constraints.

Figure 3-3 TIMES_DEMO directory

3.1.3 Description of the system represented by TIMES_DEMO

The TIMES_DEMO model is structured by regions47

. The internal regions of the TIMES_DEMO

are the Rest of the World (ROW) and the Western Europe (WEU), correspond to regions within the

area of study, and for which a Reference Energy System (RES) has been defined. Each internal

region contains processes and commodities to depict an energy system. In the RES of each region

are represented and described the following sectors:

- Supply - describing fossil fuel extraction, fossil fuel import (natural gas), renewable

potentials, materials availability (for the industry sector) and fuel transformation process

(refinery);

- Power – describing all central electricity and heat production including combined heat and

power;

- Industry - describing the industrial material flow, processes and end-uses, and

- the residential, commercial, agriculture and transportation - describing each said sector end-

uses.

For the TIMES_DEMO model the commodities exchange between the internal regions (TRADE)

are electricity and oil crude.

47

As explained in the paragraph 2.1.3.2 one needs to distinguish between external regions and internal regions.


76

3.2 How to represent in the model the main components of an energy system

This section explains how to include the desired energy system components in your model. It shows

what structure of the VEDA templates is necessary and what information has to be declared. The

input contains both structural information (the reference energy system, or connectivity of the

system elements) and technical – economic data. The TIMES_DEMO model attempts to provide an

example for each aspect of an energy system that needs to be included in most models. It follows

the recommend approach of organizing the base year system by sector, and follows the other

conventions that result in the most effective use of VEDA to manage a TIMES model. It also strives

to illustrate recommended best practice use of parameters, although VEDA and TIMES in some

cases support many variants of a parameters (e.g., efficiency can be described using EFF (for sum

of the outputs/sum of inputs of the same type, CEFF (to tie the efficiency of two commodities

directly together, or Consumption (=Input, the amount of the input commodity needed for each unit

of the output commodity).

3.2.1 How to declare regions, time horizon and time slices

The regions, time horizon and time slices, are declared in the “SysSettings” workbook, as shown in

the paragraph 2.2.3.1. The Figure 3-4 and Figure 3-5 show the declarations in the TIMES_DEMO

model, in which there are two different regions (ROW, Rest of the World and WEU, Western

Europe), twelve time periods (from 2000 to 2050) and twelve time slices (four for the seasonal level

and three for the day-night level). The book name (cell A3) must be the same name used for the B-

Y template workbook root (in this case VT_DEMOT_<sector>_V5.xls).

Figure 3-4 SysSettings: Demo model – Regions and Timeslices related declarations

Figure 3-5 SysSettings: Demo model - Time periods related declarations


77

3.2.1.1 Time slices description

The milestoneyears can be further divided in sub-annual timeslices in order to describe for the

changing electricity load within a year which may affect the required electricity generation capacity,

or other commodity flows that need to be tracked at a finer than annual resolution.

Timeslices may be organised into four hierarchy levels only: „ANNUAL‟, „SEASON‟, „WEEKLY‟

and „DAYNITE‟ defined by the internal set tslvl. The level ANNUAL consists of only one member,

the predefined timeslice „ANNUAL,‟ while the other levels may include an arbitrary number of

divisions. The desired timeslice levels are activated by the user providing entries in the SysSettings

file table ~TimeSlices. Figure 3-6 illustrates a timeslice tree, in which a year is divided into four

seasons consisting of working days and weekends, and each day is further divided into day and

night timeslices. The name of each timeslice has to be unique in order to be used later as an index in

other sets and parameters. Not all timeslice levels have to be utilized when building a timeslice tree,

for example one can skip the „WEEKLY‟ level and directly connect the seasonal timeslices with the

daynite timeslices. The duration of each timeslice is expressed as a fraction of the year by the

parameter YRFR (see the sheet Constant). The user is responsible for ensuring that each lower level

group sums up properly to its parent timeslice, as this is not verified by the pre-processor. The

definition of a timeslice tree is region-specific.

Figure 3-6 Times slices levels

In the table ~TimeSlices of the DEMO model there are 4 Seasons (S=summer, F=Fall, W=Winter

and R=Spring) and three Daynite (D = Day, N=Night and P=Peak) time slices. So constructing the

time slice tree you can obtain the following 12 time slices:

- SD (summer day), SN (summer night), SP (summer peak), FD (fall day), FN (fall night), FP (fall

peak), WD (winter day), WN (winter night), WP (winter peak), RD (spring day), RN (spring night)

and RP (spring peak).


78

3.2.1.2 Declaring Commodity Group

Commodity groups (CG) identify the collection of like input/output flows, where each commodity

corresponds to a group containing simply itself. In TIMES-VEDA, for each process the relationship

between the input commodity group and the output commodity group must be established to

properly define the efficiency ratio. As noted earlier, VEDA-FE assists the user by means of

“macro” parameters that are translated into the actual TIMES parameters needed to properly

represent this key relationship. Furthermore, it proceeds from the default premise that all processes

are output normalized with the commodity groups being all the outputs of the appropriate type (e.g.,

energy (NRG) for standard processes, demand (DEM) for demand devices) and input of the same

time (though for demand devices all NRG inputs are assumed). Yet in some cases it is

desirable/necessary to override these defaults; in this example four CGs are declared.

Figure 3-7 SysSettings: Demo model – Commodities Group

Note that in general VEDA will establish appropriate default CGs for any technology involving

multiple commodities of the same time based upon the assumption that all processes are output

normalized, where the energy commodities (NRG) determine the groups for standard processes and

the demand commodities (DEM) determine the primary commodity group for demand devices, with

the input group based upon the energy commodities.

3.2.2 How to declare commodities

In the paragraph 2.1.3.1 is described how to declare commodities. For more information you can

see "Templates Basic - Tables" at www.kanors.com/vedasupport.

3.2.2.1 General features

Commodities are assigned to different groups (Sets) according to their role in the energy system. In

the TIMES_DEMO model the following sets are used: NRG = energy carriers, DEM = demands,

ENV = environmental indicators and MAT = materials. The user must also specify the commodity

unit, and remember that the flows are measured in commodity units. The commodities and

processes of the TIMES_DEMO model are declared in the following five workbooks:

- VT_DEMOT_SUP_V5, sheet SUP_COMM;

- VT_DEMOT_ELC_V5, sheet ELC_COMM;

- VT_DEMOT_IND_V5, sheet IND_COMM;



79

- VT_DEMOT_TRA_V5, sheet TRA_COMM, and

- VT_DEMOT_OTH_V5, sheet OTH_COMM.

As you can verify in the _COMM and Processes pages of the TIMES_DEMO templates, the

declarations follow the same structure shown in Figure 2-12 for the TIMES-TUTORIAL model

(paragraph 2.2.4.2).

3.2.2.2 Emissions

In TIMES it is possible to consider any kind of emissions (e.g. CO2, NOx, SO2, SF6, etc.) using a

coefficients related to the flow of commodities or to the technologies for each sector. In the first

case, you give the combustion emission coefficients related to the consumption of fuel. . These

coefficients are applied to all the fuel consumption by the individual technologies in each sector.

For example, to track CO2 emission coefficients from six fuels (diesel oil, LPG, gasoline, natural

gas, hard coal and kerosene) the table shown in Figure 3-8 (from the VT_DEMO_SUP_V5

workbook) is provided.

Figure 3-8 Example of coefficients for combustion emissions

In the second case, when associating emission directly with specific technologies, the combustion

emission coefficients for a specific technology (only one or a group) is specified rather than

associating the emissions with the fuel consumption. The user must obviously take care if mixing

these two approaches to ensure that there is no double-counting.

If sector-based emissions are to be tracked employing the commodity flow approach, then the fuel

names will need to reflect the sector in which they are consumed (e.g., COMOILD, COMOILL,

COMOILK, COMNGA, COMCOAH) and have associated named emissions (e.g., CO2C).

In the DEMO model, in each workbook there is a sheet EMI, in which the emissions related to the

fuel consumed in that sector are declared.

3.2.2.3 Materials

In the DEMO model, the materials are declared in the MIN workbook, using mining technologies

and then are used in the industry sector (IND) as part of meeting the energy service demands.


80

3.2.3 How to declare a process: general features

Processes are assigned to different groups (Sets) according to their role in the energy system. In the

TIMES_DEMO model the following sets are used: MIN = mining process, IMP = import process,

EXP = export process, PRE = generic process, ELE = thermal electricity power plant, CHP =

combined heat and power plant, HPL = district heating plants, STGTSS (pump storage timeslices),

STGIPS (pump storage IP) and DMD = demand devices. The user must also specify the process

activity and capacity unit.

The processes of the TIMES_DEMO model are declared in the following five workbooks:

- VT_DEMOT_SUP_V5, sheet Processes;

- VT_DEMOT_ELC_V5, sheet Processes;

- VT_DEMOT_IND_V5, sheet Processes;

- VT_DEMOT_TRA_V5, sheet Processes, and

- VT_DEMOT_OTH_V5, sheet Processes.

As you can verify in the Processes pages of the TIMES_DEMO templates, the declarations follow

the same structure shown in Figure 2-13 for the TIMES-TUTORIAL model (paragraph 2.2.4.3).

3.2.3.1 Definition of process activity variables

Since TIMES distinguishes between process activity as well as commodity flows and installed

capacity; thus it is necessary to relate these types of relationship. This is done by a constraint that

equates the overall activity variable with the appropriate set of flow variables, properly weighted,

and to the total installed capacity (perhaps by vintage). This is accomplished by first identifying the

group of commodities that defines the activity (and thereby the capacity as well) of the process. In a

simple process, one consuming a single commodity and producing a single commodity, the

modeller simply chooses one of these two flows (sides) to define the activity, and thereby the

process normalization (input or output). In more complex processes, with several commodities

(perhaps of different types) as inputs and/or outputs, the definition of the activity variable requires

determination of the primary commodity group (PCG) that will serve as the activity-defining group.

For instance, the PCG may be the group of energy carriers, or the group of materials of a given

type, or the group of GHG emissions, etc. Based upon the PCG designation the like commodities on

the other side of a process as grouped into the so-called shadow primary group (SPG, which is

determined by VEDA).

As noted earlier, VEDA assists with the determination of an appropriate default PCG according to

the nature of the process. For PREs that are clarified as demand devices (DMD) it is assumed the

process is output normalized for the demand(s) services (DEM). For other PREs the processes are

again assumed to be output normalized, according to all the energy carriers (NRG) leaving the

processes. If the user wishes to input normalize a process (e.g., a refinery is usually characterized

according to the barrels of crude oil consumed) the user needs to special the PCG (e.g., CRDOIL).


81

A user may also need to define the PCG if not all of the commodities in the group define the

process activity (e.g., natural gas consumed as part of refining), these are considered associated

flows.

The topology associated with a process defined according to the commodities input and output. The

activity of a standard process is equal to the sum of one or several commodity flows on either the

input or the output side of a process as defined by the PCG. The activity of a process is limited by

the available capacity, so that the activity variable establishes a link between the installed capacity

of a process and the maximum possible commodity flows entering or leaving the process during a

year or a subdivision of a year.

3.2.3.2 Use of capacity

In each time period the model may use some or all of the installed capacity according to the

Availability Factor of that technology. For each technology, period, region and time-slice, the

activity of the technology may not exceed its available capacity, as specified by a user defined

availability factor. Thus the model may decide to use less than the available capacity during certain

time-slices, or even throughout one or more whole periods, if such a decision contributes to

minimizing the overall cost. Optionally, there is a provision for the modeller to force specific

technologies to use their capacity to their full potential or with a given number of full load hours (by

indicating that the AF is fixed (an exogenous utilization factor) rather than a maximum potential.

3.2.3.3 Defining flow relationships in a process

A process with one or more (perhaps heterogeneous) commodity flows is defined by one or more

independent input and output flow variables. In the absence of relationships between these flows,

the process would be completely undetermined, i.e. its outputs would be independent from its

inputs. We therefore need one or more relationships stating that the ratio of the sum of some of its

output flows to the sum of some of its input flows is equal to a constant (which is akin to an

efficiency). In the case of a single commodity in, and a single commodity out of a process, this

equation defines the traditional efficiency of the process (see the Electricity Power Plants, section

3.2.4.3). With several commodities, this constraint may leave some freedom for individual output

(or input) flows, as long as their sum is in fixed proportion to the sum of input (or output) flows (see

the Flexible Refinery, section 3.2.4.2). An important rule for this constraint is that each sum must

be taken over commodities of the same type (i.e. in the same group, say: energy carriers, or

emissions, etc.). In TIMES-VEDA, for each process the relationship between the input commodity

group and the output commodity group must be established to properly define the efficiency ratio.

As noted earlier, VEDA-FE assists the user by means of “macro” parameters that are translated into

the actual TIMES parameters needed to properly represent this key relationship. The various

VEDA-FE efficiency related parameters are briefly described here:


82

- EFF - can be used for defining process efficiencies of most processes, the parameter value

defines the amount of activity that can be produced by one unit of flow of a commodity or

commodities on the shadow side of the process;

- CEFF (commodity specific efficiency) - if the efficiency of the power plant is different

depending upon the fuel consumed, and

- INPUT (or Consumption) – indicating the amount of a commodity needed for a unit of

output. Input and output attributes can be conveniently used for specifying fixed relations

between the process activity and individual input or output flows that are not part of the

primary group (PG), or even between two flows (e.g., emissions that are commodity

dependent).

As the efficiency of the power plant is different depending upon the fuel consumed it is necessary to

use the CEFF (commodity specific efficiency).

3.2.3.4 Limiting flow shares in flexible processes

When either of the commodity groups, input or output, contains more than one element, the

previous relationship allows a lot of freedom on the values of flows. The process is therefore quite

flexible. To limit this flexibility within acceptable performance ranges, the share of each flow

within its own group may be controlled within ranges48

. For instance, a refinery output might

consist of four refined products with 5% losses. The user may then want to limit the flexibility of

the slate of outputs with four flow shares, as shown in the section 3.2.4.2, by specifying the

maximum shares for the various outputs. The commodity group being subjected to shares may be

on the input or output side of the process.

3.2.3.5 Peaking Reserve Requirements (time-sliced commodities only)

In TIMES it is required that the total capacity of all processes producing a commodity at each time

period and in each region must exceed the average demand in the time-slice when the highest

demand occurs by a certain percentage. This percentage is the Peak Reserve Factor,

COM_PKRSV, and is chosen to insure against several contingencies, such as possible commodity

shortfall due to uncertainty regarding its supply (e.g. water availability in a reservoir), unplanned

equipment down time, and random peak demand that exceeds the average demand during the time-

slice when the peak occurs. This constraint is therefore akin to a safety margin to protect against

random events not explicitly represented in the model. In a typical cold country the peaking time-

slice for electricity (and natural gas) will be in the Winter (Winter-Peak if a peak timeslice is

defined), and the total electric plant generating capacity (or gas supply) must exceed the Winter-

Peak demand load by a certain percentage. In a warm country the peaking time-slice may be in the

48

It is always advisable to leave at least one commodity in a group unconstrained to avoid over specify the process.


83

Summer for electricity (due to heavy air conditioning demand). The peaking constraint is created

for each time-slice of a time-slice commodity49

.

3.2.4 How to declare specific processes

3.2.4.1 Mining process and import/export processes

The mining and import/export processes are described in the TIMES-TUTORIAL paragraph 2.2.5.1

and paragraph 2.2.5.2. The user can see how these processes are described in the TIMES-DEMO in

the workbook VT_ DEMOT_SUP_V5, sheets MIN and IMP-EXP. As can be seen there are two oil

crude and two hard coal domestic production options and one natural gas import option.

3.2.4.2 Flexible refinery

A simplified flexible refinery is an energy process (PRE) that consumes one commodity (Comm-

IN) and produces four (or more) commodities (Comm-OUT) in a flexible manner, using the

parameter Share~UP (in principle is also possible to use FX or LO) to control the individual flows

as related to the total output. As reflected in Figure 3-9 over the period each output commodity can

be at most equal the fifty percent of the total production from the refinery.

The flexible refinery must be described using a Primary Commodity Group (PCG = OILCRD, for

example) in order to define the activity of the process (SSCDRFLX00) as the sum of the four output

flows (see Figure 3-9). In this way is possible to normalize the output, thus the efficiency (1.05) is

being defined from output flows group OILCRD (sum of all output flows) to crude oil input, and

represents a 5% loss across the process.

In addition to the activity of a process, one has to define the relationship of the activity to the

capacity unit of the process. This is done using the capacity-to-activity parameter (Cap2Act),

applied to the primary commodity group. In the example in Figure 3-9 the capacity of the refinery

process is defined in PJ/a, while activity is measured in PJ. Thus, the conversion factor is 1 (a

capacity of 1 PJ/a produces in a year an activity of 1PJ) and can be omitted. By default the Cap2Act

is equal to one.

The last parameter, LIFE, is used to describe the operational lifetime of the refinery in terms of the

number of years. By default the economic life (payback period) is assumed to equal the operational

life, though the user could distinguish them (using ELIFE), if desired.

In Figure 3-9 an example of a flexible refinery take from the TIMES-DEMO (workbook

VT_DEMOT_SUP_V5 sheet Refinery) is shown. Besides those parameters already discussed a tax

49

The user may control for which time-slices the peaking constraints are generated, by explicitly identifying said time-

slices, but as default a peaking constraint is created for all time-slices at the commodity time-slice level.


84

is applied to some of the output commodities (via FLO_TAX) and sector CO2 are tracked based

upon the consumption of the crude oil.

Figure 3-9 Flexible Refinery

3.2.4.3 Electric power plants

The electricity power plants are special process that generate electricity (ELE), consuming one or

more commodities and producing electricity by time-slice.

In Figure 3-10 three examples are shown:

- a fossil fuel plant that consumes coal (COAHAR);

- a fossil fuel plant that has the flexibility to consume oil and/or natural gas (OILDST and

GASNAT), and

- a seasonal reservoir hydro-electric power plant that consumes a renewable commodity

(HYDRO) to produce electricity.

The parameters used to describe the electricity power plants in the Figure 3-10 define the:

- topology (Comm-IN/OUT);

- efficiency (EFF or CEFF);

- all the availability factor of a technology are50

:

- NCAP_AF - availability factor relating a unit of production (process activity) in

timeslice s to the current installed capacity.

- NCAP_AFA - annual availability factor relating the annual activity of a process to the

installed capacity. Provided when „ANNUAL‟ level process operation is to be

controlled.

- NCAP_AFS - availability factor relating the activity of a process in a time slice

(SEASON/WEEKLY/DAYNITE) being at or above the process time slice level to the

installed capacity. If for example the process time slice level is „DAYNITE‟ and

NCAP_AFS is specified for time slices on the „SEASONAL‟ level, the sum of the

„DAYNITE‟ activities within a season are restricted, but not the „DAYNITE‟ activities

directly.

50

For more details see TIMES documentation Part II, table 12.


85

- NCAP_AFC - can be used to describe commodity specific availability factors. It is thus

meaningful for only processes with several commodities in the primary group.

NCAP_AFC is automatically combined with any normal annual availability factors

defined for the process. If no normal availability factors have been defined on the

timeslice level tslvl, an upper limit of 1 is always used as the default. If NCAP_AFC is

specified for only some but not all commodities in the primary group, a default value of

1 will be used for any missing commodities. NCAP_AFC can be used for any

processes, including storage and trade processes. In the case of storage processes, only

the output flow of commodity COM is considered in the availability constraint.

- NCAP_AFAC, is just a shorthand alias for NCAP_AFC(…,ANNUAL).

- the contribution of the power plant towards meeting the peak requirement (PEAK).

- Investment along with fixed and variable operating and maintenance costs (NCAP_COST,

FIXOM and VAROM)51

;

- capacity to activity conversion factor (Cap2Act,by default equal to 1);

- level of the existing installed capacity in the base year (STOCK);

- the flow share (SHARE~<bd>) establishing the relationship for the commodities available

to the dual-fuelled power plant, and

- emission coefficients (ENVACT~<env> for emissions other than CO2.

The first power plant (ECOASTM000) is described using the commodity input, the commodity

output, the overall process efficiency (EFF), the annual availability factor (AF), the

capacity/activity transformation (CAP2ACT), the existing stock by region

(STOCK~<regionname>), a lower production bound (BNDACT~LO), an emission coefficient on

the output (ENVACT~<emission commodity name>) and the peak parameter (PEAK).

The peak parameter (never larger than 1) specifies the fraction of a technology capacity that is

considered to be secure and thus will most likely be available to contribute to the peak load in the

highest time-slice of a year for a commodity (electricity or heat only); many types of supply

processes can be regarded as predictably available with their entire capacity contributing during the

peak and thus have a peak coefficient equal to 1, whereas others (such as wind turbines or solar

plants in the case of electricity) are attributed a peak coefficient less than 1, since they are on

average only fractionally available at peak (e.g., a wind turbine typically has a peak coefficient of

0.25 or 0.3).

In addition to the activity of the power plant, one has to define the relationship to the capacity unit

of the process. This is done using the parameter Cap2Act, applied to the primary commodity group.

In this example the capacity of the power plants is defined in GW. Since the capacity and activity

units are different (GW for the capacity and PJ for the activity), the user has to supply the

conversion factor from the energy unit embedded in the capacity unit to the activity unit. This

conversion factor is 31.536 PJ/GW.

51

Although for simplicity FIXOM and VAROM are omitted for these plants.


86

Figure 3-10 Electricity power plants

The second power plant (EGASOIL000) needs to relate the two input commodities to the process in

various ways. As the efficiency of the power plant is different depending upon the fuel consumed it

is necessary to use the CEFF (commodity specific efficiency, see section 3.2.3.3) rather than the

overall process efficiency parameter EFF, and provided a flow share to reduce the freedom of the

model (as explained in the section 3.2.3.4). For EGASOIL000 a lower flow share on input (Share-

I~LO) is assigned for the commodity OILDST, requiring that at least the 20% of the fuel consumed

is distillate.

With regard to the hydro-electric power plant An efficiency of 0.33 is a conventional value used,

representing the estimated fossil equivalent that would be necessary if this plant was not used), and

is also valid for the wind technologies, other renewable power plants, and for nuclear power plants.

This is the minimum sets of parameters to describe an electricity power plant, some other efficiency

and advanced parameters are described in VEDA-FE interface under Tools, Supported Attributes.

The emissions in the TIMES-DEMO are related to the commodity consumption, as described in the

TIMES-TUTORIAL, but it is also possible to assign the emissions directly to the technology (see

section 2.2.5.4). This is especially important for pollutants for which the emission coefficients

depend on the individual technologies. Some other examples are shown in the TIMES-DEMO

(workbook VT_DEMOT_ELC_V5 sheet ELC).

For other power plant examples see the workbook VT_DEMO_ELC_V5 sheet ELC.

3.2.4.4 Cogeneration power plant

Cogeneration power plants or combined heat and power plants (CHP) are plants that consume one

or more commodities and produce two commodities, electricity and heat. One can distinguish two

different types of cogeneration power plants according to the flexibility of the outputs, a back

pressure process and a condensing process, as shown in Figure 3-14.


87

A typical cogeneration power plant includes many of the same parameters mentioned in section

3.2.4.3 for electric generation plants52

as well as:

- PEAK - the cogeneration power plants are technologies that work on the DAYNITE

timeslice level, thus the peak parameter must be related to this timeslice level. For example

is possible to relate the peak to the winter-peak timeslice (WP, as in the Figure 3-14) or the

summer-peak (SP);

- CHPR (FX, LO or UP) is defined to be the heat-to-power ratio, and

- the ratio of electricity lost to heat gained (CEH).

Back pressure turbine

Back pressure turbines are system in which the ratio of the production of electricity and heat is

fixed, the electricity generation is directly proportional to the steam.

Figure 3-11 Back pressure turbine characteristic curve

In a real system a back pressure turbine is defined using the electrical efficiency (ETAelBP), the

thermal efficiency (ETAthBP), and the load utilization. Thus in TIMES-VEDA, a back pressure

system is characterised as follows.

ETAel = 45%, ETAth = 30%, Load utilization period = 3500 h/a in TIMES-VEDA become

EFF = ETAelBP + ETAthBP = 75%

NCAP_AFA = load utilization period/duration of a year =3500/8760 = 0.40

CHPR~FX = ETAthBP/ETAelBP = 0.67

CEH = 1

If the CHPR parameter is fixed (FX) the production of electricity and heat is in a fixed proportion,

but one could also use a (LO) CHPR for defining the back-pressure point, if so desired (to allow by-

passing the turbine to produce more heat). CEH can be either 0 (or missing) or 1 for fixed back

pressure mode processes. If it is zero, the activity represents the electricity generation and the

52

Many other parameters can be given.


88

capacity represents the electrical capacity; if it is 1, the activity represents the total energy output

and the capacity represents the total capacity (power+heat). The comment about the fixed CHPR is

basically correct, but one could also use a 'LO' CHPR for defining the BP point, if so desired (to

allow by-passing the turbine to produce more heat).

It is also possible to describe a back pressure CHP process by means of the CEFF parameter. In this

case the equations shown are still valid. The first technology shown in Figure 3-14 is an example of

back pressure cogeneration power plant. The others parameters are the same as those described for

the electricity power plant.

Condensing combined heat and power plant

The condensing pass-out or extraction turbine version of a CHP process does not have to produce

heat, permitting only electricity to be generated, and permitting the amount of heat generated to be

directly adjusted to the heat demand, where the electricity generation is reciprocally proportional to

heat generation (electricity losses because of heat extraction).

Figure 3-12 Condensing combined heat and power characteristic curve

Pass-out or extraction turbines are thus described slightly differently.

1. Electricity to heat ratio, via parameter CEH such that:

a) <= 1: electricity loss per unit of heat gained (moving from condensing to backpressure

mode), indicating that activity is measured in terms of electricity, or

b) >= 1: heat loss per unit of electricity gained (moving from backpressure to condensing

mode), indicating that activity is measured in terms of total output (electricity plus heat).

2. Efficiencies, according to 1:

a) are specified for the condensing point, or

b) are specified for backpressure point.

3. Costs, according to 1:

a) are specified based according to condensing mode, or

b) are specified based on total electricity and heat output at backpressure point.


89

4. Electric loss per unit of heat gained (CHPR): Ratio of heat to power at backpressure point; at

least a maximum value is required, but in addition also a minimum value may be specified

Figure 3-13 Condensing combined heat and line fuel

Thus in TIMES-VEDA, a condensing system is characterised as follows.

ETAel = 54% at the condensing point; ETAelBP = 45%; ETAthBP = 44%

Load utilization periodheat = 3500 h/a

Load utilization periodelectricity = 7500 h/a in TIMES-VEDA become:

EFF = ETAel = 0.54

CHPR~UP = ETAthBP/ ETAelBP = 0.98

VDA_CEH = (ETAel – ETAelBP)/ETAthBP = 0.2045

AFAC~ELC = 7500/8760 = 0.856

AFAC~HEAT = 3500/8760 = 0.40

Also for the condensing CHP processes it is possible to use the CEFF parameter. In this case the

equation shown remains valid.

The first technology shown in Figure 3-14 is an example of a back pressure cogeneration power

plant and its parameters and the second one is an example of a condensing cogeneration power

plant. Besides the CHP specific parameters other parameters related to power plants may be used.

This is the minimum sets of parameters need to describe a CHP, other relevant parameters are

described in VEDA-FE interface under Tools, Supported Attributes.

Figure 3-14 Cogeneration power plants


90

3.2.4.5 District heating plants

The district heating plants are heat generation technologies (HPL) that consume one or more

commodities and produce one time-sliced commodity (heat).

In Figure 3-15 is shown an example of district heating plants, the parameters used to describe this

process are basically the same parameters mentioned in section 3.2.4.3 for electric generation

plants, except that the output commodity is heat and the capacity unit PJ/a, and thus the Cap2Act is

1.

Figure 3-15 District heating plants

3.2.4.6 Cars, trains, bus – converting activity, capacity, demand units; and dual-purpose devices

In this section various transport technologies are described (cars, train, buses, etc.). Generally these

technologies are defined as demand technologies (DMD). This means that they consume an energy

commodity to produce directly the energy services demanded.

The parameters to describe a demand technology are:

- topology Comm-IN/OUT;

- efficiency (EFF or CEFF);

- availability factor (AF);

- commodity dependent availability factor (upper limit) (AFAC);

- conversion factor Cap2Act;

- cost (VAROM);

- installed capacity in the base year (STOCK), and

- process activity to commodity flow (ACTFLO).

It might be the case that the unit in which the commodity(ies) of the primary commodity group are

measured is different from the activity unit. An example is shown in Figure 3-16. The activity of the

transport technology passenger car (CAR) is defined by the two demands they service (cars short

distance – CAR_SD and cars long distance - CAR_LD), which are measured in million of

passenger kilometres (MPKms). The activity of the process is, however, defined in million vehicle

kilometres (MVKms), while the capacity of the process CAR is defined as thousand units of cars


91

(„000 units). Thus the conversion factor from capacity to activity (Cap2Act) needs to describe the

average mileage of a car per year. The process parameter (ACTFLO) contains the conversion

factor from the activity unit to the commodity unit of the primary commodity group PCG). In this

example the factor corresponds to the average number of persons per car (1.5). The availability

factor (AF) describes the maximum kilometers available for a car in a year. In addition a

commodity dependent availability factor (upper limit, AFAC) corresponding to the maximum kms

related to short/long distance relationship relative to the overall AF parameter. In this example the

AF is 20000 and the AFAC~CAR_SD is 0.8, this means that the maximum availability of the car

for short distance travel (CAR_SD) is 80% of the overall usage and is thus equal to 20000*0.8 =

16000.

The same parameters used for the cars can also characterize trains, buses, and so on for other multi-

use demand technologies. Other examples of transportation processes are available in the

VT_DEMOT_TRA_V5 workbook.

Figure 3-16 Cars

3.2.4.7 Car process- defining demand and the load shape

The demand devices are used to satisfy an energy service demand. For example cars are used to

satisfy private transport demand on short and long distance. In VEDA is necessary to define a table

with base year energy service demand (see Figure 3-17). Then at the column level is defined the

demand by region. The load curve describes seasonal, weekly and diurnal variations in demands

(externally given as inputs). In this example is shows the transport cars demand, with seasons and

diurnal variations, described in the TIMES_DEMO with a variable load curve, as shows in Figure

3-17. In figure Figure 3-17 a "Table level declarations" is used. The Table level declaration to

define load curve is ~FI_T: COM_FR. This table level declaration use COM_FR as the attribute for

all values in the table that don't have an attribute specification at the column or row level.

Figure 3-17 Cars demand and load shape


92

3.2.4.8 Industrial process

In this an example of the industry chain for the Iron and Steel sector is described. In this case only

the last technology (e.g. finishing process) is described like a demand technology, where the other

technologies of the chain are described as (upstream) processes in the chain. This means that they

consume energy commodities and/or materials to produce new materials useful for the Iron and

Steel chain production. The last process, that is a demand technology, finally consumes energy

commodities and materials produced in the chain to satisfy the iron and steel demand.

The parameters used to describe the industrial processes in the Figure 3-10 are similar to those with

some variations being the way the alternative commodities are handled:

- consumption of fuels/materials (INPUT), and

- production of materials (OUTPUT).

The first step of the iron and steel production chain is the iron and steel pellet and sinter production,

as shows in Figure 3-18. The materials produced in the first step are then used to produce raw iron,

as shows in Figure 3-19. The third step is related to the crude steel production and characterized

from the consumption of fuels and materials produced in the second step, as shows in Figure 3-20.

Figure 3-18 Industrial processes – Iron and steel pellet and sinter production

Figure 3-19 Industrial processes – Raw iron production

Figure 3-20 Industrial processes – Crude steel production

The last step is the iron and steel production, through a finishing process described as demand

technology, as shows in Figure 3-21.


93

Figure 3-21 Industrial processes – Iron and Steel production

3.2.5 How to construct a scenario file

The Scen_<appl>.xls files contain additional information and parameters for the entire RES,

commodities and technologies (rule-based). The important thing to understand about scenario files

is that they can only manipulate information associated with previously declared RES components,

and that new commodities and technologies may not be added via scenario files, though parameters

may be. For this example there are two files:

To construct a scenario file, from VEDA-FE Navigator click „NEW‟ in the Scenario Files window

an put name (e.g. test) in the pop-up window. An excel file with the scenario file structure will be

open, as shows in Figure 3-22.

Figure 3-22 New Scenario file from VEDA-FE Navigator

To define subsets of technologies/commodities a set of headers (see Table 3.1) are used under

which the user specifies masks using text/wildcards (“?” for a single character, * for any number) to

identify the qualifying technologies. Technology qualifiers identify candidates based upon set

members (Pset_Set), and/or masks for topology (commodity in/out, Cset_CI/O), and/or

name/description masks (Pset_PN/D). In general if nothing appears below a specification column

the values provided for the parameter apply to all entries. Exclude is done by “-“<mask>. Multiple

masks may be specified separated by “,”.


94

Table 3.1: Headers in the scenario files to define subsets of technologies/commodities

Header Description

Region Region name

PSet_Set Process set. Comma separated list of process names allowed (wild cards allowed)

PSet_PN Process name set. Comma separated list of process names allowed (wild cards allowed)

PSet_PD Process description set. Comma separated list of process description allowed (wild

cards allowed)

Pset_CI Process commodity input set. Comma separated list of input commodities to define a set

of processes allowed (wild cards allowed)

Pset_CO Process commodity output set. Comma separated list of output commodities to define a

set of processes allowed (wild cards allowed)

CSet_Set Commodity set. Comma separated list of process names allowed (wild cards allowed)

CSet_CN Commodity name set. Comma separated list of commodity names allowed (wild cards

allowed)

CSet_CD Commodity description set. Comma separated list of commodity descriptions allowed

(wild cards allowed)

To filling-in this sheet with the parameter, it‟s enough a to check from VEDA-FE interface, Tools

and Supported Attribute.

Figure 3-23 shows an example of scenario file for the Demo. This is a scenario file to impose a CO2

taxation on the Green House Gas (GHG). In the column attribute there is the parameter

COM_TAXNET, below the column year there is the year in which to apply the taxation, the value

is in the column WEU and the commodity name to which apply the taxation is below Cset_CN

(Commodity name of the commodity set).

Figure 3-23 Scenario file from the Demo

3.2.6 How to construct User Constraints (UC)

While TIMES provides most all the core equations that are needed to properly represent the

individual components of the energy system (e.g., commodity balance, process operation, capital

stock turnover), the user will almost always needs to introduce additional constraints to adequately

represent aspects of their particular energy system. Such constraints often involve clusters of

technologies (e.g., solar, or wind, or hydro each may be subject to a total potential installed

capacity, renewable portfolio (RPS) or CAFE standards may require that a group of technologies

meet a certain percentage of overall electricity generation or average vehicle fleet efficiency by

vehicle class).


95

To enable this necessary, and powerful, flexibility User Constraints (UC) may be built. TIMES

User Constraints allow relationships to be established between most any of the TIMES model

variables, (summed over region, period, time-slice as desired) as well as input parameters. Here a

couple of practical examples are described as implemented for TIMES_DEMO.

The majority of User Constraints fall into two general categories, absolute and share. The role of

absolute user constraints is to control the investment, capacity or operation of a set of processes in

absolute terms. Examples include:

- Electricity generation should consume at least 400 PJ of gas in each period;

- Maximum hydro potential is 50MW in 2005 and remains constant from 2010 on at 2010,

see Figure 3-21;

- Geothermal should produce at least 3 TWh per year by 2020, and 10 by 2050, and

- The total nuclear capacity should be at most 4 GW by year 2020, and at most 10 by year

2050.

The role of share User Constraints are to control the investment, capacity or operation of a set of

processes (subset) as the share of a larger set (BigSet). Examples include:

- A maximum of .5PJ and a minimum of .05PJ of coal may be consumed for the generation of

electricity, while the maximum share from gas is 90%, see Figure 3-21;

- At least 5% of electricity generation should be wind based by 2020;

- At least 10% of the residential space heating should be based on natural gas;

- Small cars may take at most 30% of the automobile travel demand, and

- At least 60% of residential lighting will use the conventional incandescent bulbs.

3.2.6.1 User Constraints (UC) in Scenario Files

In the TIMES_DEMO the user constraints are in the Scen_UCTest workbook. How to construct the

UC is defined in User Constraints at www.kanors.com/vedasupport. Declaring UC Sets:

3.2.6.2 How to construct special flow share scenarios

As discussed above, the key to defining a share constraint is to identify two sets of technologies: a

BigSet (BS), and a SubSet (SS). For example, in the first case, BS would be all technologies that

produce residential heat for existing rural houses, and the SS would comprise those among the BS

that consume wood. Another case could be the use of renewable hydro power plant in the electricity

production, in which the BS would be all technologies that produce electricity, and SS would

comprise those among the BS that consume hydro “fuel”.

In principle, the BS and SS can be defined in many ways, but a vast majority of cases can be

handled by defining the BS based on commodity produced (CP), and the SS based on commodity

consumed (CC). For TIMES VEDA makes it possible to construct such relationships by means of a

special flo-share scenario (AFS scenario).



96

The Figure 3-24 shows the UConstraints section in the VEDA-FE Navigator. It lists all the

UConstraints scenarios stored in the UConstraints folder in SuppXLS folder of the Template path

and allows to create new UC Scenarios and import the existing ones. Use it to create a new User

Constraint share scenario by:

1. Clicking on New button in the UConstraints section of VEDA-Naviator to create a new UC

Scenario, Figure 3-24;

2. Entering the scenario name in the prompt box (should be unique across all scenarios and the

name given can be of max 20 chars long; the Scenario is saved with the name given by the

user preceded by “ScenUC_” in UConstraints folder in SuppXLS folder of the template

path);

3. Selecting the GDX file associated with a previously run Base scenario from which the base

year fuel shares are to be extracted;

4. After the GDX file is read, a confirmation box appears that asks if an existing UC file‟s

transformation information needs to be inherited: if yes clicked, another open file box

appears indicating the UC file to be inherited here (in this case the years for UC

specification are taken from the source file), while if no clicked, another prompt box appears

asking the years for which UC‟s are to be specified, as a comma separated string which

greater than the base-year.

- TheUC scenario file is then created, with or without the inherited transformation, where

if inherited the complex definitions are also written in the new UC file.

5. Then double click on the created file (in the Figure 3-24 Test_AFS_UC) to open the UC File

for edit (see below), and

6. Click SYNC to import the UC File to associated Database (Adratio).

3.2.6.3 The UC scenario file contains the following sheets:

1. BaseYrDataValues<baseyr> - that shows the data values of each CP, CC combination in all

regions in the base year (filled by VEDA);

2. BaseYrShares<baseyr> - that shows the data shares of each CP, CC combination in all

regions in the base year (filled by VEDA);

3. A transformation sheet Trans<yr> (where the user provides the operations to be applied to

the base year shares over time, and

4. A Final Data Shares Sheet Final<yr> for each year chosen for UC specification (created by

VEDA).

When a file is edited via VEDA-Navigator mode:

- The user cannot modify the BaseYrDataValues and the Final Sheets at all, and

- New CP, CC combinations can be added to the BaseYrShares sheet. [As a new CP, CC

combination is added to the sheet the corresponding shares for all the regions are

evaluated immediately. The CP and CC can have comma separated values, can include

wildcards viz. * %? Underscore <_> is not recognized as a wildcard but a literal.]


97

Figure 3-24 User Constraint Scenario in VEDA Navigator

3.2.7 How to construct a new technologies file (SubRes)

3.2.7.1 How to construct a SubRes file

The folder „SubRes_TMPL‟ contains rule-based templates with data specification (file SubRES_B-

NewTechs.xls) and transformation (for the respective SubRES) for new technologies;

The parameters to describe a new technology are the same shown in the previous section for the

base year technologies. Generally speaking however, new technologies will not have any existing

STOCK, but will have two additional parameters:

- START, to describe the first period from which the new technology is available in the

model, and

- INVCOST, to assign an investment cost to the new technology.

The following figures show some example of new technologies described in the SubRes.

In the SubRes all the commodities (including emissions) used to describe the new technologies

must be declared, regardless whether they have already been declared in the base year templates.

Figure 3-25 Example of new electricity power plant in the SubRes


98

Figure 3-26 Example of new industry technologies in the SubRes

For other examples is possible to see the SubRES_B-NewTechs template in the SubRES_TMPL

folder.

3.2.7.2 How to construct a SubRes transformation file

See Templates Basic - files n SubRes section at www.kanors.com/vedasupport.

3.2.8 How to construct a demand file

The folder „SuppXLS‟ contains the sub-folder „Demand‟ where the demand driver (DEMO_Base)

and sensitivity series and region-segment driver allocation table (Dem_Alloc+Series), applied over

time to produce the base year projection. Figure 3-27 shows a driver table (~DRVR_Table) with the

drivers (GDP, Population, etc.) for the two regions (ROW and WEU), with their initial allocation

for each region.

Figure 3-27 Demand driver

The next two figures show the sensitivity series (~Series, Figure 3-28) and the drivers allocation

(DRVR_Allocation), Figure 3-29) table. A new series may be added by means of a new row with

the name of the series and the new values for each year.



99

Figure 3-28 Demand sensitiviy series

Figure 3-29 shows the driver allocations for each of the regions‟ demands. Each demand is

associated with a driver, along with a calibration and sensitivity series. The calibration and

sensitivity column are the series described in the ~Series table, Figure 3-29.

Figure 3-29 Demand drivers allocation table

3.2.9 How to construct a trade scenario

3.2.9.1 How to declare a trade matrix

To exchange commodities between the regions53

it is necessary to declare a trade matrix and some

attributes for the associated inter-region exchange process (IRE). This is accomplished by means of

the Trade scenario file (ScenTrade_TRADEAttribs) with parameters for the inter-regional trade

technologies.

The TIMES_DEMO model exchanges electricity (ELC) and oil crude (OILCRD) between the two

regions (ROW and WEU). The associated trade matrix is declared in an Excel file, as shows in the

Figure 3-30.

53

To know more about the TIMES trade structures see the “Documentation for TIMES model – PART II” from page

215 available at http://www.etsap.org/Docs/TIMESDoc-Details.pdf

http://www.etsap.org/Docs/TIMESDoc-Details.pdf


100

Figure 3-30 Trade matrix declaration

3.2.9.2 How to declare a trade parameters

To declare the trade parameters, is possible to construct a scenario file, as shown in Figure 3-31

(from the Demo).

Figure 3-31 Trade parameters

3.3 Interpolation and extrapolation

Time-dependent user input parameters are specified for specific years, the so called data years.

These data years do not have to coincide with the model years needed for the current run. Reasons

for differences between these two sets are for example that the period definition for the model has

been altered after having provided the initial set of input data leading to different milestone years or

that statistical data are only available for certain years that do not match the model years. In order to

avoid burdening the user with the cumbersome adjustment of the input data to the model years, an

inter/extrapolation routine is embedded in the TIMES model generator and in the VEDA software.

The inter/extrapolation routine distinguishes between a default inter/extrapolation that is

automatically applied to the input data and an enhanced user controlled inter/extrapolation that

allows the user to specify inter/extrapolation rules for each time series explicitly. Independent of the

default or user-controlled inter/extrapolation options, TIMES inter/ extrapolates (using the standard

algorithm) all cost parameters in the objective function to the individual years of the model as part

of calculating the annual cost details.


101

3.3.1 Defaults inter/extrapolation

The standard default inter/extrapolation method interpolates linearly between data points, while it

extrapolates the first/last data point constantly backward/forward. The parameters given in Error!

Reference source not found.Table 3.2 are by default NOT inter/extrapolated in this standard

default method. All other parameters are by default both interpolated and extrapolated in the default

method.

In many cases VEDA-FE provides defaults that often meet the user‟s needs and leave the user to

focus of the actual data.

Table 3.2: Parameters not being inter/extrapolated by default

Parameter Justification Alternative

default method

ACT_BND

Bounds, may be intended at specific periods

only

Migration

CAP_BND

NCAP_BND

FLO_FR

FLO_SHAR

STGOUT_BND

STGIN_BND

COM_BNDNET

COM_BNDPRD

COM_CUMNET

COM_CUMPRD

COM_CHRBND

IRE_BND

IRE_XBND

UC_RHST User constraints may be intended for specific

periods only

Migration

UC_RHSRT

UC_RHSRTS

NCAP_AFM Interpolation is meaningless for these

parameters (parameter value is a discrete

number indicating which MULTI curve

should be used).

None

NCAP_EFFM

NCAP_FOMM

NCAP_FSUBM

NCAP_FTAXM

NCAP_AFX Interpolation is meaningless for these

parameters (parameter value is a discrete

number indicating which SHAPE curve

should be used).

None

NCAP_EFFX

NCAP_FOMX

NCAP_FSUBX

NCAP_FTAXX

NCAP_PASTI Parameter describes past investments for

individual vintage years so is not

interpolated.

None

NCAP_PASTY Parameter describes number of years over

which to distribute past investments. None

COM_BLVAL Blending parameters at the moment are not

interpolated. None

PEAKDA_BL

COM_BPRICE Base prices for elastic demands are obtained

from baseline solution

None

CM_MAXCO2C Bound may be intended at specific years only None


102

As shown in Table 3.2, for bound and RHS parameters an alternative default method of

interpolation/extrapolation is applied: migration. Migration means that data points are interpolated

and extrapolated within each period but not across periods. This method thus migrates any data

point specified for other than milestoneyr years to the corresponding milestoneyr within the period,

so that it will be effective in that period.

3.3.2 Enhanced user-controlled interpolation / extrapolation

The inter/extrapolation facility provides maximum flexibility by allowing the user to control the

interpolation of each time series separately. Many bounding constraints as well as market and

product allocation constraints might be applicable either to only specific years or to the continuous

times pan of the full time horizon, or to a subset thereof. The possibility of controlling interpolation

on a time series basis improves the independence between the years found in the primary database

and the data actually used in the individual runs of a TIMES model. In this way the model is made

more flexible with respect to running scenarios with arbitrary model years and period lengths, while

using basically the very same input database.

The enhanced interpolation/extrapolation facility provides the user with options to control the

interpolation and extrapolation of each individual time series (Table 2). The option 0 does not

change the default behaviour. The specific options that correspond to the default methods are 3 (the

standard default) and 10 (alternative default method for bounds and RHS parameters). Non default

interpolation/extrapolation can be requested for any parameter by providing an additional instance

of the parameter with an indicator in the YEAR index and a value corresponding to one of the

integer valued Option Codes (see Table 2 and example below).

This control specification activates the interpolation/extrapolation rule for the time series, and is

distinguished from actual time series data by providing a special control label (“0”) in the YEAR

index. The particular interpolation rule to apply is a function of the Option Code assigned to the

control record for the parameter. Note that for log linear interpolation the Option Code indicates the

year from which the interpolation is switched from standard to log linear mode. TIMES user shell(s)

will provide mechanisms for imbedding the control label and setting the Option Code through

easily understandable selections from a user friendly dropdown list, making the specification simple

and transparent to the user.

The enhanced interpolation/extrapolation facility provides the user with the following options to

control the interpolation and extrapolation of each individual time series.

- Interpolation and extrapolation of data in the default way as predefined in TIMES. This

option does not require any explicit action from the user.

- No interpolation or extrapolation of data (only valid for non cost parameters).

- Interpolation between data points but no extrapolation (useful for many bounds). See option

codes 1 and 11 in Table 3.3 below.


103

- Interpolation between data points entered, and filling in all points outside the interpolation

window with the EPS value. This is useful for e.g. the RHS of equality type user constraints,

or limitations on future investment in a particular instance of a technology, which should

often have a continuous value of EPS to be effective. See option codes 2 and 12 in Table 2

below.

- Forced interpolation and extrapolation throughout the time horizon. Can be useful for

parameters that are by default not interpolated. See option codes 3, 4, and 5 as well as 14

and 15 in Table 3.3 below.

- Log linear interpolation beyond a specified data year, and both forward and backward

extrapolation outside the interpolation window. Log linear interpolation is guided by relative

coefficients of annual change instead of absolute data values.

Table 3.3: Option codes for the control of data interpolation

Action Option code Applies to

Default interpolation/extrapolation (see above) 0 (or none) All

No interpolation/extrapolation < 0 All

Interpolation but not extrapolation 1 All

Interpolation, but extrapolation with EPS 2 All

Full interpolation and extrapolation 3 All

Interpolation and backward extrapolation 4 All

Interpolation and forward extrapolation 5 Bounds, RHS

Migrated interpolation/extrapolation within periods 10 Bounds, RHS

Interpolation migrated at end-points, no extrapolation 11 Bounds, RHS

Interpolation migrated at ends, extrapolation with EPS 12 Bounds, RHS

Interpolation migrated at end, backward extrapolation 14 Bounds, RHS

Interpolation migrated at start, forward extrapolation 15 Bounds, RHS

Log-linear interpolation beyond YEAR YEAR (≥1000) All

Apart from the migrating options 10–15, all the other enhanced interpolation options described

above are available for all TIMES parameters. The migrating options are available for all bound and

RHS parameters, which are listed in Table 1 above (excluding CM_MAXCO2C, for which

migration is of no use because the parameter is effective for any given year). Note that because

option 10 is the default method for bound and RHS parameters, and it is not available for other

parameters, there is no need to ever use this option explicitly. It is mentioned in Table 2 for

completeness only.

3.3.3 Interpolation of cost parameters

As a general rule, all cost parameters in TIMES are densely interpolated and extrapolated. This

means that the parameters will have a value for every single year within the range of years they

apply, and the changes in costs over years will thus be accurately taken into account in the objective

function. The user can use the interpolation options 1–5 for even cost parameters. Whenever an


104

option is specified for a cost parameter, it will be first sparsely interpolated/extrapolated according

to the user option over the union of modelyear and datayear, and any remaining empty data points

are filled with the EPS value. The EPS values will ensure that despite the subsequent dense

interpolation the effect of user option will be preserved to the extent possible. However, one should

note that due to dense interpolation, the effects of the user options will inevitably be smoothed.


105

4 Appendix A - Getting Started with Problem – Defining and

Describing the Area of Study54

This appendix is a primer on energy systems analysis. Approaching energy as a system instead of a

set of elements gives the advantage of identifying the most important substitution options that are

linked to the system as a whole and cannot be understood looking at a single technology or

commodity or sector.

Systems analysis applies systems principles to aid decision-makers in problems of identifying,

quantifying, and controlling a system. While taking into account multiple objectives, constraints,

resources, it aims to specify possible course of action, together with their risks, costs and benefits.

After an excursus on the peculiarities of energy as a system, this appendix illustrates how to

proceed to the three steps of the analyses: identification, quantification and control. It is intended to

illustrate how complex energy related matters are, it hints to the complexity of decision-making in

energy related matters and it shows why it helps using ETSAP Tools to represent energy systems

and compile alternative development scenarios.

4.1 The multiple dimensions of energy systems

Present energy systems are the result of complex country dependent, multi-sector developments.

Although each decision in this n-step path may have provided rational answers based upon energy,

engineering, economic or environmental reasons (for short: 4E), it is hard to find rationality in the

overall system. Furthermore, decisions take into account several other important dimensions that,

broadly speaking, are part of humanities or social sciences.

The four main dimensions encompassed by energy systems analyses will be shortly illustrated in

the following pages. Although social sciences touch aspects as fundamental as the four mentioned

above, they are not treated here because the ETSAP modelling tools are not yet in the position to

represent them explicitly and quantitatively. At most, ETSAP tools help measuring how large is the

gap between actual and theoretical systems.

54

Appendix drafted by G.C. Tosato


106

4.1.1 Energy: from primary resource to end-use services 55

An energy system comprises an energy supply sector and energy end-use. The energy supply sector

consists of a sequence of elaborate and complex processes for extracting energy resources,

converting these into more desirable and suitable forms of energy, and delivering energy to places

where the demand exists. The end-use part of the energy system provides energy services such as

cooking, illumination, comfortable indoor climate, refrigerated storage, transportation, and

consumer goods. The purpose, therefore, of the energy system is the fulfilment of demand for

energy services.

The architecture of an energy system can be represented by a sequential series of linked stages,

alternating commodities and processes, connecting various energy conversion and transformation

processes that ultimately result in the provision of goods and services (see Figure 4-1). A number of

examples are given for energy extraction, treatment, conversion, distribution, end-use (final

energy), and energy services in the energy system. The technical means by which each stage is

realized have evolved over time, providing a mosaic of past evolution and future options.

Primary energy is the energy that is embodied in resources as they exist in nature: the chemical

energy embodied in fossil fuels (coal, oil, and natural gas) or biomass, the potential energy of a

water reservoir, the electromagnetic energy of solar radiation, and the energy released in nuclear

reactions. For the most part, primary energy is not used directly but is first converted and

transformed into electricity and fuels such as gasoline, jet fuel, heating oil, or charcoal. Primary

energy is expressed in common units of PJ (international standard) or Tonnes oil equivalent

(frequently used)56

.

Final energy is the energy transported and distributed to the point of final use. Examples include

gasoline at the service station, electricity at the socket, or fuel wood in the barn. All final energy

vectors are expressed in the common energy unit (GJ), but it is clear to all users that the same GJ

content of wood fuels is different from electricity: the latter can produce work directly, the former

has to undergo the Carnot cycle and losses before producing work.

The next energy transformation is the conversion of final energy in useful energy, basically heat and

work, by means of energy end- use devices, such as boilers, engines or motor drives. Useful energy

is measured at the crankshaft of an automobile engine or an industrial electric motor, by the heat of

a household radiator or an industrial boiler, or by the luminosity of a light bulb. In principle useful

55

This paragraph and the following one are taken from N. Nakichenovich et al., CLIMATE CHANGE 1995, Impacts,

Adaptation and Mitigation of Climate Change: Scientific – Technical Analyses, Contribution of Working Group II to

the Second Assessment Report of the Intergovernmental Panel for Climate Change, WMO – UNEP, Cambridge

University Press, 1996, 878 pages, Chapter B., pages 75-78 56

In other words, primary energy consumption is an abstract concepts, calculated by converted the different forms of

energy into a common unit. The conversion can be calculated at the “physical content” equivalent (adopted by IEA), or

at the “substitution principle” (adopted by EIA). The former method converts electricity from nuclear at 33% efficiency,

geothermal at 10% efficiency and all other non-biomass renewable sources at 100% efficiency. The latter converts

every non fossil / non biomass electricity / heat at the average efficiency of existing fossil power plants. According to

the editor of this report, this second method is more appropriate for economic evaluation (it uses a concept similar to the

marginal value) and gives equal weight to each kWh, produced by wither nuclear or renewables.


107

energy may be expressed in common energy units, but in practice it is used in sector or application

specific energy related units (thermie, lumen, etc.).

In conjunction with non-energy end-use devices, useful energy provides energy services, such as

moving vehicles, warm rooms, process heat, or light. Energy services are expressed in specific

units, such as passengers or tons-kilometre, square-meters of heated flats, tons of cement, and even

value added or labour force in economic producing sectors.

Figure 4-1: The energy system: schematic diagram with some illustrative examples of the energy

sector and energy end-use and services.


108

Energy services are the result of a combination of various technologies, infrastructures (capital),

labour (know-how), materials, and energy carriers. Clearly, all these input factors carry a price tag

and, within each category, are in part substitutable for one another. From the consumer's

perspective, the important issues are quantity, quality and cost of energy services. It often matters

little what the energy carrier or the source of that carrier is. It is fair to say that most consumers are

often unaware of the "upstream" activities of the energy system. The energy system is service-

driven (i.e., from the bottom up), whereas energy flows are driven by resource availability and

conversion processes (from the top down). Energy flows and driving forces interact intimately.

Therefore, the energy sector cannot be analysed in isolation: It is not sufficient to consider only how

energy is supplied; the analysis also must include how and for what purposes energy is used.

In 1990, 385 EJ of primary energy produced 279 EJ of final energy delivered to consumers,

resulting in an estimated 112 EJ of useful energy after conversion in end-use devices. The delivery

of 112 EJ of useful energy left 273 EJ of rejected energy. Most rejected energy is released into the

environment as low-temperature heat, with the exception of some losses and wastes such as the

incomplete combustion of fuels.

4.1.2 Engineering: technology efficiency and system efficiency

Energy is conserved in every conversion process or device. It can neither be created nor destroyed,

but it can be converted from one form into another. This is the first law of thermodynamics. For

example, energy in the form of electricity entering an electric motor results in the desired output-

say, kinetic energy of the rotating shaft to do work - and losses in the form of heat as the undesired

by-product caused by electric resistance, magnetic losses, friction, and other imperfections of actual

devices. The energy entering a process equals the energy exiting. Energy efficiency is defined as the

ratio of the desired (usable) energy output to the energy input. In the electric- motor example, this is

the ratio of the shaft power to the energy input electricity. In the case of natural gas for home

heating, energy efficiency is the ratio of heat energy supplied to the home to the energy of the

natural gas entering the furnace. This definition of energy efficiency is sometimes called first-law

efficiency.

A more efficient provision of satisfying energy services not only reduces the amount of primary

energy required but, in general, also reduces adverse environmental impacts. Although efficiency is

an important determinant of the performance of the energy system, it is not the only one. In the

example of a home furnace, other considerations include investment, operating costs, lifetime, peak

power, ease of installation and operation, and other technical and economic factors. For entire

energy systems, other considerations include regional resource endowments, conversion

technologies, geography, information, time, prices, investment finance, operating costs, age of

infrastructures, and know-how.

The overall efficiency of an energy system depends on the individual process efficiencies, the

structure of energy supply and conversion sector, and the energy end-use patterns. It is the result of

compounding the efficiencies of the whole chain of energy supply, conversion, distribution, and


109

end-use processes. The weakest link in the analysis of the efficiency of various energy chains is the

determination of energy services and their quantification, mostly due to the lack of data about end-

use services and actual patterns of their use.

In 1990, the global efficiency of converting primary energy sources to final energy forms, including

electricity, was about 72%. The efficiency of converting final energy forms into useful energy is

lower, with an estimated global average of 40%. The resulting average global efficiency of

converting primary energy to useful energy, then, is the product of the above two efficiencies, or

29%. Because detailed statistics for most energy services do not exist and many rough estimates

enter the efficiency calculations, the overall efficiency of primary energy to services reported in the

literature spans a wide range, from 15 to 30%.

Figure 4-2: Major energy and carbon flows through the global energy systems in 1990

How much energy is needed for a particular energy service? The answer to this question is not so

straightforward. It depends on the type and quality of the desired energy service; the type of


110

conversion technology; the fuel, including the way the fuel is supplied; and the surroundings,

infrastructures, and organizations that provide the energy service. Initially, energy-efficiency

improvements can be achieved in many instances without elaborate analysis through common

sense, good housekeeping, and leak-plugging practices. Obviously, energy service efficiencies

improve as a result of sealing leaking window frames or installing a more efficient furnace. If the

service is transportation - getting to and from work, for example - using a transit bus jointly with

other commuters is more energy- efficient than taking individual automobiles. After the easiest

improvements have been made, however, the analysis must go far beyond energy accounting.

Here the concept that something may get lost or destroyed in every energy device or transformation

process is useful. For instance, in terms of energy, 1kWh of electricity and the heat contained in 43

kg of 20°C water are equal (i.e., 3.6 MJ). At ambient conditions, it is obvious that 1kWh electricity

has a much larger potential to do work (e.g., to turn a shaft or to provide light) than the 43 kg of

20°C water. Another, more technical, example should help clarify the difference. Furnaces used to

heat buildings are typically 70 to 80% efficient, with the latest, best-performing condensing

furnaces operating at efficiencies greater than 90%. This may suggest that little energy savings

should be possible, considering the high first-law efficiencies of furnaces. Such a conclusion is

incorrect. The quoted efficiency is based on the specific process being used to operate the furnace -

combustion of fossil fuel to produce heat. Because the combustion temperatures in a furnace are

significantly higher than those desired for the energy service of space heating, the service is not

well matched to the source, and the result is an inefficient application of the device and fuel. Rather

than focusing on the efficiency of a given technique for the provision of the energy service of space

heating, one needs to investigate the theoretical limits of the efficiency of supplying heat to a

building based on the actual temperature regime between the desired room temperature and the heat

supplied by a technology.

The ratio of theoretical minimum energy consumption for a particular task to the actual energy

consumption for the same task is called second-law (of thermodynamics) efficiency57

. Consider

another example: Providing a temperature of 27°C to a building while the outdoor temperature is

2°C requires a theoretical minimum of about one unit of energy input for every 12 units of heat

energy delivered to the indoors (according to the second law of thermodynamics). To provide 12

units of heat with an 80% efficient furnace, however, requires 12/0.8, or 15, units of heat. The

corresponding second-law efficiency is the ratio of ideal to actual energy use (i.e., 1/15 or 7%). The

first-law efficiency of 80% gives the misleading impression that only modest improvements are

possible. The second law efficiency of 7% says that a 15-fold reduction in final heating energy is

theoretically possible. For example, instead of combusting a fossil fuel, Goldemberg et at. (1988)

give the example of a heat pump, which extracts heat from a local environment (outdoor air, indoor

exhaust air, ground-water) and delivers it into the building. A heat pump operating on electricity

can supply 12 units of heat for 3 to 4 units of electrical energy. The second-law efficiency improves

to 25-33% for this particular task- still considerably below the theoretical maximum efficiency. Not

accounted for in this example, however, are energy losses during electricity generation. Assuming a

57

The concept of second-low efficiency opens the door to the use of the concept of exergy.


111

modern gas-fired, combined- cycle power plant with 50% efficiency, the overall efficiency gain is

still a factor of two compared with a gas furnace heating system. In practice, theoretical maxima

cannot be achieved. More realistic improvement potentials might be in the range of half of the

theoretical limit. In addition, further improvements in the efficiency of supplying services are

possible by task changes – for instance, reducing the thermal heat losses of the building to be heated

via better insulated walls and windows.

There are many difficulties and definitional ambiguities involved in estimating the efficiencies

according to the second thermodynamic principle for comprehensive energy source-to-service

chains or entire energy systems. The analysis of individual conversion devices is comparatively

simpler than the analysis of energy systems efficiencies to useful energy or even to energy services.

All indicate that primary-to-service (second-law) efficiencies are as low as a few percent. An

overall primary-to-useful energy second law efficiency of 21% has been estimated for Japan, less

than 15% for Italy, 32% for Brazil, which reduces to 23% when primary energy to service second

law efficiency are estimated (2.5% for the United States). Other estimates of global and regional

primary-to-service energy second law efficiencies vary from ten to as low as a few percent.

The theoretical potential for efficiency improvements is very large; current energy systems are

nowhere close to the maximum levels suggested by the second law of thermodynamics. However,

the full realization of this potential is impossible. Friction, resistance, and similar losses never can

be totally avoided. In addition, there are numerous barriers and inertias to be overcome, such as

social behaviour, vintage structures, financing of capital costs, lack of information and know-how,

and insufficient policy incentives.

The principal advantage of second-law efficiency is that it relates actual efficiency to the theoretical

(ideal) maximum. Although this theoretical maximum can never be reached, low efficiencies

identify those areas with the largest potentials for efficiency improvement. For fossil fuels, this

suggests the areas that also have the highest emission-mitigation potentials.

4.1.3 Economics: the value of energy systems

Since the industrial revolution took place, the economic development as a whole is powered by

energy and the global 2005 GDP of about 55 TUS$200058

would not be possible. However this

economic development and the associated welfare for the people come at a cost that is considerable,

directly and indirectly. What part of the world economic resources are consumed in order to supply

producers and consumers with the energy services they need?

The direct cost of providing primary fuels has been of the order of 1.5 TUS$ globally in 2000. In

the year 2000, with an average spot price of crude oil of 26.8$/bbl, this cost represented about 4%

of the global GDP at market prices (about 37 TUS$). In 2005, with a yearly average spot price of

38.1 $/bbl, this share is much higher. But at current future prices of crude oil approaching 100 $/bbl

58

The figure is taken from the IEA “Key Energy Statistics”, page 48, at purchase power parity.


112

(November 2007) the marginal value of primary energy supply approaches 10 percent over the

GDP. The additional annual cost of generating electricity, fuels and heat adds another considerable

amount. If annual costs of transmission and distribution are added, the total cost of supplying to the

economic producers and the families the amount of energy demanded is on the order of 15% of the

global GDP. Slightly more than half of this cost is borne by families, the rest by industries.

However, the economic weight of energy systems as a whole is much higher. Actually, if the energy

system efficiency concept explained above is taken into account, the economic weight of the system

affected by energy policies ranges between 35 to 50%. In fact it includes all end use devices that

transform final energy into useful energy and into the energy services demanded by final users – i.e.

motor and engines, heating systems and thermal insulation, industrial boilers and ovens, etc. Their

energy efficiency improvement potential is much higher than in the primary supply sectors. In a

system analysis view, this part of the system is even more important than the supply side when it

comes to controlling the future development of the system.

4.1.4 Emissions and the environment

The damage to the environment is the major indirect cost caused by present energy systems59

.

Substances emitted into the atmosphere by energy technologies such as:

- power plants;

- refineries;

- incinerators;

- factories;

- domestic households;

- cars and other vehicles;

- animals and humans;

- fossil fuel extraction and production sites;

- offices and public buildings, and

- distribution pipelines

are mainly responsible for:

- global warming/climate change;

- acidification;

- air quality degradation, and

- damage and soiling of buildings and other structures.

“Carbon dioxide is the most important anthropogenic greenhouse gas”60

. “The primary source of the

increased atmospheric concentration of carbon dioxide since the pre-industrial period results from

59

The level of this indirect cost is highly debated. Methodologies used so far to assess this value (for instance ExternE

and its extension elaborated in NEEDS – both are EC projects) have some degree of reliability in the evaluation of the

physical impacts but diverge when it comes to converting the damage into monetary units because most damaged goods

are public goods, whose values are not given by the market but have to be assessed from subjective evaluations. 60

Quotes from the Summary for Policymakers of the IPCC Fourth Assessment Report “Climate Change 2007: The

Physical Science Basis” (February 2007)


113

fossil fuel use, …” (about 85%). “The carbon dioxide radiative forcing increased by 20% from 1995

to 2005, the largest change for any decade in at least the last 200 years.” The radiative forcing

contribution of CO2 equals now the total net anthropogenic contribution to radiative forcing (1.66

W/m2).

The acidification of the soil is due to the air emissions of sulphur oxides, nitrogen oxides and partly

ammonia. Over 90% of sulphur and nitrogen oxides are emitted from energy systems61

.

Local pollution, mainly urban, originates from anomalous concentrations of carbon monoxide,

volatile organic compounds, particulate matter, sulphur and nitrogen oxides. About 90% of carbon

monoxide emissions originate from energy systems (mainly transport) and about 60% of volatile

organic compounds.

Decoupling the benefits of using energy from its disadvantages so far has been partly successful

only in the case of sulphur oxides, with some improvements in for nitrogen oxides, volatile organic

compounds and particulate matters. Decoupling carbon dioxide emissions from the use of energy

remains the major environmental issue of energy systems analyses.

4.2 The systems analysis approach: identification of the areas of study

The identification of the system is the first step towards its formal representation in models and

their use for carrying out mental experiments aimed at exploring how the system might evolve

under different circumstances and how it is possible to control it. The most important elements to be

identified when an energy system is approached for analysis seem:

- Scope of the analyses;

- Boundaries;

- Time frames;

- Components (Elements, Parts), and

- Connections, interdependencies and chains.

A short illustration of each element follows.

4.2.1 Scope of the analysis

Before starting the analysis of an energy system the following general elements have to be

identified:

- Who is the client?

- What is the aim, what are the specific objectives?

- Who will conduct the analysis?

61

You can access to more information from the EMEP/CORINAIR Emission Inventory Guidebook – European

Environment agency (http://www.eea.europa.eu/).

http://www.eea.europa.eu/


114

- Are the necessary skills available?

- What budget has been allocated for the study?

Establishing these elements can guide one through conflicting requirements of the analysis, for

instance between strategic and operative planning, among different level of comprehensiveness and

detail.

Operative and strategic planning is distinguished by the time span considered and by other factors

related to the energy, technological and socio-economic framework. Operative planning looks at

short-term optimisation from minutes to days, and an otherwise fixed technical energy system and

socio-economic framework. Strategic planning tries to include long-term technological and socio-

economic developments.

Operative planning requires a large amount of detail within the sub-systems, because high accuracy

is required. In strategic planning, too much detail in the sub-systems can often obstruct the view of

the total system behaviour for long-term developments. Instead it is more important to consider the

interdependencies between the large numbers of sub-systems. A planner cannot control too many

details at the same time, since he has only limited resources to collect the necessary data and to

build and exercise a model, aimed at helping him to understand the complex interdependencies.

Additionally there are restrictions to the ability of the tools to handle complexity and to process the

data required for very large models in reasonable time. As a result, operative planning is carried out

on a sub-system level with a limited time horizon and little consideration of comprehensive aspects.

Strategic planning, on the other hand, is done in a comprehensive analysis with a long time horizon

and less detail on the sub-system's level. Despite the different characteristics and purposes of

operative and strategic planning, insights obtained from one of the model groups can be used to

better describe the behaviour of the other model, e.g. relationships obtained from the operative

planning model can be included in a simplified and aggregated form in the strategic planning

model.

Sub-system and comprehensive analyses are distinguished by the extent of their system boundaries.

Sub-system analysis is restricted to a limited number of sub-systems within the whole technical

energy system. Concentrating only on a sub-system offers the possibility to study the

interdependencies of the system in much detail, but influences from or to other subs-ystems are

considered only in a simplified manner. A comprehensive analysis, on the other hand, tries to treat

all important sub-systems and their interdependencies within one model. In this framework part of

the details and comprehensiveness of actual systems can be retained, part lost. The expertise of the

analyst helps reaching a balance between the details of an actual system and the synthesis of any

mental representations.


115

4.2.2 Boundaries

Geographical boundaries are defined by the nature of the problem: energy R&D problems or

climate change mitigation studies may require a global model, security issues require at least a

regional approach, designing taxes and subsidies requires a national system, waste disposal options

or local pollution problems are to be studied at the urban / local level. A general description of the

system should cover the geographic situation, climate, temperature, the population (historical

series), and the differing conditions between rural and urban areas, the economy, and the main

features of the energy system.

Sometimes the identification is more complex, particularly at the local level. When there are huge

exchanges between the geographical system and the surrounding areas – think to vehicle traffic of

commuters to a town or the trade of energy intensive materials of a country – it is sometimes better

to define the boundaries of the „logical system.‟

It is also to be decided whether a single comprehensive energy system has to be considered or it is

to be sub-divided into more than one region. In the first case the representation will be based upon

an aggregate energy balance and will describe an average situation. If more sub-regions are to be

employed to describe for instance the European Union, such as the Northern, Central and Southern

parts, or each member state, it will necessary to identify and quantify energy flows, technologies,

emissions and economic values separately for each region, as well as the possibilities for trade

between the regions.

4.2.3 The time dimension

Statistics and macro variables refer to annual values. Keeping the year as a base, the time dimension

of the system can be explored in two dimensions62

. One looks at the variations from one year to

another (inter-years), the second at the variations inside the typical year (intra-annual). The system

control looks at long term developments along the years; other important energy systems aspects,

such as electricity and heat, or gas supply and traffic, refer to intra-annual dimensions.

4.2.3.1 Time horizon

The main concern of energy systems analyses and the rationale for building energy system models

is the study of possible future developments control policies. This is reflected in the time horizon of

the analyses: it spans from years (short term) to decades (medium term) to a century (long term) or

more as it happens now in relation to climate changes mitigation policies impact analyses. But

62

In fact the time development of the system along the years can be viewed from two different perspectives, similarly to

fluid dynamics. The point of view adopted here is Eulerian in the sense that the time development of the system as a

whole is followed year after year. An alternative approach is adopted by Life Cycle Analyses, where the time

development of each element of the system is followed along the years from cradle to grave.


116

understanding possible future developments implies studying past behaviour and reconstructing the

present layout in an appropriate (maximum) level of detail.

Year by year data are required by the most detailed analyses, such as for local energy environment

planning and analyses often focused more on the operation of than the investments in the system.

However, when the time horizon is far way in the future, it is customary to follow the time

development of the system by time periods of variable lengths, yearly in the short term, every five

years in the medium, and every twenty or more years in the long period. More periods offer the

possibility to evaluate the system, and thus adjust decisions and strategies, more frequently, but also

increase the size of the problem and the amount of data to be processed.

Analysing the intersection between energy and climate change mitigation issues requires the

adoption of a very long-term perspective. Energy infrastructure takes a very long time to build and

has a useful life often measured in decades. New energy technologies take time to develop and even

longer to reach their maximum market share. Similarly, the impact of increasing concentrations of

greenhouse gases from human activities develops over a very long period (from decades to

centuries), while policy responses to climate change threats may only yield effects after

considerable delay. Analysis that seeks to tackle these issues must take a similarly long term view –

looking ahead at least thirty to fifty years.

4.2.3.2 Time granularity

The problem of intra-annual time analyses is easily understandable by looking at the most important

energy commodity of any system, electricity. Electricity is accessible everywhere in developed

countries and satisfies the demand for many energy services of consumers. Since this demand is

different across countries – for instance space heating and cooking is provided by electricity in

some countries and not in others – the primary energy weight of electricity varies from about 50%

of Sweden and France to 27% in China. Also the daily, weekly and yearly profile of electric

demand is different across countries. In fact electricity is not a single commodity as there are as

many electricity markets (and prices) as there are hours in a year.

A load curve shows how the demand for electricity, heat or cooling, etc. varies over time. An

example for district heating production is shown in Figure 4-3. This load curve is made up of daily

averages, from Jan. 1 to Dec. 31. The diagram shows the typical pattern of high energy production

during the winter and the small energy production during the summer for water heating.

The energy demand variation is an indication of the heating power demand, i.e. the load curve

contains information on both power and energy demand. The load curve is often shown in a

simplified form with load levels presented as monthly averages, Figure 4-4. This makes the figure

easier to read and it also makes calculations based on the load curve more practical to perform.

However, a lot of information can be lost, e.g. about peak load.


117

Figure 4-3: - Example of load curve profile for district heating production 63

Figure 4-4: Monthly average load curve for district heating production

Another way of representing the changing demand for energy is the load duration curve. The

duration curve associated with the previous district heating example is shown in Figure 4-5. This

diagram is made up of the data from the detailed load curve. The duration curve starts with the day

having the highest energy production and ends with the day having the lowest energy production. In

the duration curve it is possible to see how many days during the year the energy production was

higher than 600 MW. The more detailed the data used for the development of the duration curve,

the better and ”smoother” the curve will be. It is therefore not a good idea to simplify the duration

curve with monthly average load data, since this would result in a less useful duration curve.

63

This figure, as well as the other two of this paragraph, are taken from the ALEP Guide Book (www.etsap.org)


118

The load curve includes more information than the duration curve, since it contains the information

regarding at what point during the year a specific load level occurs. This information is lost when

the load curve is transformed into a duration curve. If, for example, both heating and cooling

demands in an analysed energy system are known, and if they are in some way interconnected, then

the load curve will show that the peak level of heating occurs in January, whereas the peak load of

cooling occurs in July, i.e. never peaking during the same season. Such information is impossible to

extract from a duration curve.

The load curve is typically used in order to calculate which production plants should be operating to

cover the district heating load, day by day. The duration curve is suitable for more principle

consideration of base load / peak load plants, and calculating the total yearly energy production

from each type of plant given their capacities.

Figure 4-5: Duration curve for district heating production

The balance between detail and dimension of the problem is problematic in this case. While the

dispatching problems require hour by hour analyses (sometime even at shorter time intervals),

systems analyses for very long term problems slice the year in few intervals by season, weekday

and hour, in order to grasp the essential points of the problem64

.

64

In the modelling practice, 3 or 6 time slices allow a first order analysis of peak vs. off-peak supply problems, 16-20

time slices appear necessary to analyse in some detail the different dynamics over the years of demands for energy

services and the possible contribution of renewable energy sources. The different consumption and supply patterns are

then reflected in the TIMES parameter COM_FR.


119

4.2.4 Components

In this context there are two main components in the system: commodities and technologies.

4.2.4.1 Commodities

All energy goods that flow in the system are included in this category. This spans from primary

sources such as coal or crude oil to final energy vectors, such as gasoline, heat or electricity. All

these goods are normally measured in energy units. The category extends to include all energy

services that the consumers needs, such as mobility – measured in passengers*kilometre – space

heat – measured for instance in square meters per degree*days – or the material that are produced

using considerable amounts of energy, such as steel, cement or polyethylene.

This category also encompasses commodities such as air emissions (e.g., carbon, sulphur and

nitrogen oxides), or wastes to be disposed (such as spent nuclear fuels or municipal wastes)65

.

How much of the actual system has to be comprehended in the identification effort depends on the

objective of the analyses. Sometimes, the analysis not only includes all energy commodities for

which statistical data are available, but it also extends to detailed energy services in all producing

and consuming sectors. For instance in the transport sector the analysis often requires to split

demands (and end-use devices) among Private Passenger Road Car Short Distance, Private

Passenger Road Car Long Distance, Public Passenger Road Bus Urban, Road Freight, and Rail

Freight.

4.2.4.2 Technologies

Technologies are the second essential element of the system. They are even more important than

commodities if the system has a high technical content, as an energy system does. In this context,

energy supply processes include power plants and refineries, and transmission / distribution

infrastructures, such as pipelines and electric grids, are upstream technologies. But in the same

category are included all devices that use the energy supplied by energy plants and infrastructures.

Therefore refrigerators, cooking ranges, vehicles, kilns and boilers are demand technologies.

In general an analysis includes all identifiable stages that transfer energy from outside into the

system and, inside the system from one energy good/service to another. In this category it is not

easy to determine the level of detail that balances the need to represent the actual nature of the

system and that to keep the representation of the system a manageable and digestible size. Usually

hundreds of power plants are squeezed into a few dozens of classes, millions of refrigerators into

few types, etc.

65

Sometimes the “undesired” commodities are included in categories different from the “demanded”. The latter have a

positive price, the former a negative price, determined by the equilibrium point between the level of demage “supplied”

by the system and accepted by the consumers. In some cases it is more useful to revert the sign of the “negative”

commodities and model the service of reducing noxious emissions or wastes.


120

4.2.5 Connections: the Reference Energy System

The components, commodities and technology, are then represented as a Reference Energy System

(RES), a network representation of all of the technical activities required to supply various forms of

energy to satisfy end-use activities, see Figure 5.6. Analytical formulations are described to

examine all operations involving specific fuels including their extraction, refinement, conversion,

transport, distribution, and utilization. Each of these activities is represented by a box (process) in

the network for which efficiency, environmental impact, and cost coefficients may be specified. The

processes are connected by links which represent the commodity. The network is quantified for a

given year by the level of the energy demands, the energy flows through the supply activities (all

associated technologies in a chain) that are required to serve those demands. The RES begins from

the existing capacity in place, calibrated for the 1st year commodity flows.

In addition, the RES contains technologies that, are known (or speculated) but not yet available on

the market, or due to too high costs or lack of reliability are not competitive today, but may be

commercial in the future. The Reference Energy System thus is a schema used to represent the full

structure of the energy system being studied, at an appropriate level of detail.

The RES describes the flow of energy from the sources to the final use. It shows all flows of energy

from the primary energy supply, large scale and small scale energy conversion, different

distribution forms and the final use of energy in different sectors. Additionally the RES usually

contains useful information on energy demand and even energy services (see Figure 4-6). It may

also show associated emissions at each point in the network. The RES, however, is not a

geographical representation of the energy system, except to the extent that multiple regions may be

represented.

Using the RES it is possible to see how energy flows and how energy conversion technologies

influence the fuel-technology chains in an energy system. This means, that the benefit of coupled

production can be estimated according to its contribution to both the district heating sub-system and

the electrical sub-system. The roles of these sub-systems can be evaluated from the perspective of

the entire energy system and the requirements on this system. This overall perspective is

particularly important when one evaluates demand side energy conservation technologies, i.e., the

balance between supply and conservation measures, or the cost-efficiency of a proposed investment

to control emissions.

While the RES is a graph of all relevant energy flows within the energy system, an energy balance

contains the values of all energy flows. These flows can be included on a RES diagram or presented

in separate tables. The RES may contain more conversion levels like distribution, end use

technologies and useful energy demands, which are normally not included in an energy balance.

The RES is preferably built-up according to certain practical recommendations:

1) Sources and primary energy supply - the RES begins at the far left of the diagram with the

input flows of energy, e.g. oil, natural gas, coal, petrol and imported electricity;


121

Figure 4-6: Example of RES – Reference Energy System


122

2) Transformation processes follow next to modify the fuels forms, e.g. oil refining and

preparation of pellets from biomass [In many cases it is more natural to describe e.g. refined

oil or biomass pellets as the primary energy supply (or available energy carriers), since

processing may have taken place outside of the studied geographic area and had no

noticeable influence on its energy system.];

3) Electricity and heat plants - next energy enters the large energy conversion technologies, e.g.

electricity production plants, district heating plants and combined heat and power plants

(CHP), to produce electricity and low-temperature heat;

4) Distribution systems - for the large scale conversion plants follow for different energy

forms, e.g. electricity, district heating and natural gas;

5) End use technologies - are the small scale energy technologies that consume final energy to

meet the demand for energy services, e.g. oil fired boilers for multi-family houses, solar

heating systems for single family houses, electrical appliances, petrol fuel cars and small

scale combined heat and power plants, as well as conservation measures which reduce the

need for certain energy services, and

6) Useful energy demand – appears on the right-most part of the RES as the energy services

needed for different kinds of applications, e.g. space heating, lighting and cooking.

While not shown on the RES examples here, it is often advisable for the analyst to pass

commodities through “dummy” sector processes to adjust the name of commodities according to

the consuming sector (e.g., COMELC, RSDDSL, CONCOA, etc.) to facilitate construction of the

energy balance tables, and make it easier to track commodities use in general.

Figure 4-7 depicts a small portion of a hypothetical RES containing a single energy service demand,

namely residential space heating.

Moving from right to left, there are three end-use space heating technologies using the gas,

electricity, and heating oil energy carriers (commodities) respectively. These energy carriers in turn

are produced by other technologies, represented in the diagram by one gas plant, three electricity-

generating plants (gas fired, coal fired, oil fired), and one oil refinery. To complete the production

chain on the primary energy side, the diagram also represents an extraction source for natural gas,

one for coal, and two for crude oil (one extracted domestically and then transported by pipeline, and

the other one imported). This simple RES has a total of 13 commodities and 13 processes. These

elements form a topology where a simple rule is imposed that in the RES every time a commodity

enters/leaves a process (via a particular flow) its name is changed (e.g., wet gas becomes dry gas,

crude becomes pipeline crude). This simple rule enables the inter-connections between the

processes to be properly maintained throughout the network without the need to explicitly assign

numbered interconnections.

The specific parameters for declaring the RES, or „topology,‟ to the TIMES model generator is the

Comm-IN/OUT designation associated with each technology. That is the topology is defined by the

commodities entering and leaving each process.. The VEDA users‟ interface presents sections of

the RES to assist the analyst with visualizing the RES at any stage.


123

Gas

extraction

Coal

extraction

Oil

extraction

Oil

Import

Gas fired

Power plant

Coal fired

Power plant

Oil fired

Power plant

Oil

refinery

Gas

Plant

PipelineOil

Furnace

Electric

Heater

Gas

Furnace

Ga

s in

gro

und

Co

al in

gro

und

Oil

in g

rou

nd

Impo

rted

Oil

De

live

red

Cru

de

Dry

Ga

s

Co

al

Wet G

as

Cru

de o

il

LF

O

HF

O

Ele

ctr

icity

Ho

me

spa

ce

He

atin

g

Gas

extraction

Coal

extraction

Oil

extraction

Oil

Import

Gas fired

Power plant

Coal fired

Power plant

Oil fired

Power plant

Oil

refinery

Gas

Plant

PipelineOil

Furnace

Electric

Heater

Gas

Furnace

Ga

s in

gro

und

Co

al in

gro

und

Oil

in g

rou

nd

Impo

rted

Oil

De

live

red

Cru

de

Dry

Ga

s

Co

al

Wet G

as

Cru

de o

il

LF

O

HF

O

Ele

ctr

icity

Ho

me

spa

ce

He

atin

g

Figure 4-7 Partial view of a simple Reference Energy System

4.3 The systems analysis approach: quantification

Once the system and its main components are identified, the estimate and quantification step can

start. Depending on the availability of statistical and technical information, the identification of the

system can be revised. On top of the usual difficulty of gleaning base information in sectors that

sometimes protect their data for competition problems, the analytical problem here is to harmonise

the four aspects in a consistent framework. This often shows that not all data sources are reliable

and additional assumptions / corrections are necessary.

4.3.1 Flows of energy commodities

The energy flows are described in the national balance (see Table 4.1). The balance contains the

energy consumption data for the base-year (the fuel consumption by end-use) and thereby holds the

basic structure of the model. Generally the balance contains data on the total fuel consumption in

each aggregate sector (commercial, industrial, residential, transport, and perhaps agriculture).


124

Table 4.1: Global summary energy balance for 1973 [IEA, 2004] (a)

1973, in Mtoe (+) Coal Crude

oil

Oil

Prod.

Nat.

Gas Nuc. Hydro Bio.

Elec.*,

Other TOTAL

Indigenous Prod. 1476 2936 - 994 52 319 673 9 6459

Imports 140 1577 410 73 - - 0 8 2209

Exports -130 -1611 -441 -73 - - 0 -8 -2264

Stock Changes 12 -22 -16 -15 - - 0 - -41

TPES 1498 2880 -47 980 52 319 673 9 6363

Intl. Marine Bunkers - - -119 - - - - - -119

Transfers - -43 49 - - - - - 5

Statistical Diff. 0 12 -7 5 - - - 0 9

Electricity Plants -557 -23 -318 -159 -52 -319 -3 499 -932

CHP Plants -88 - -28 -51 0 - -1 101 -67

Heat Plants -9 - -1 -1 - - -1 7 -5

Gas Works -9 -1 -9 14 - - - - -5

Pet. Refineries - -2800 2773 - - - - - -28

Coal Transf. -169 1 -3 0 - - 0 - -171

Liquefaction Plants -2 0 - - - - - - -1

Other Transf. - 4 -5 0 - - -11 - -13

Own Use -34 -3 -162 -107 - - 0 -58 -364

Distribution Losses -7 -7 0 -8 - - - -43 -65

TFC*** 622 21 2121 672 - - 657 516 4608

Industry Sector 358 16 556 381 - - 99& 277 1686

Transport Sector 33 - 905 18 - - - 10 966

Other Sectors 226 - 528 273 - - 559& 228 1814

Non-Energy Use 5 4 132 - - - - 141

& split assumed by the editor; *: Other includes: solar, wind, geothermal in the primary section, electricity and heat in

the use section. + The Total Final Consumption is equal to the original IEA data; the Total Primary Energy Supply is

not, because here all non-fossil electricity has been converted to primary equivalent by using the same average fossil

efficiency of 38.79%. When the physical energy principle is used – as the IEA statistical office does – nuclear weighs

three times more than hydro, although its electric output is nearly the same. (a) The methodology of energy balances is

well explained in chapter 7 of the “Energy Statistics Manual” issued by the statistical office of the International Energy

Agency, downloadable from http://www.iea.org/textbase/nppdf/free/2005/statistics_manual.pdf.

Only in industry might national energy balances split consumption by sub-sector. However, it is

necessary to be able to distinguish the nature of the demands for energy services, for instance in the

residential sector, limiting the analysis of consumption to a single sector is not satisfactory because

it does not distinguish between completely different services and end use devices, such as those

necessary for lighting or space heating or cooling.

The disaggregation is possible through the collection of additional data from other sources and their

integration in the national energy balance. Sometimes expert judgements assumptions are needed to

disaggregate the fuel consumption down to sub-sector and end-use levels. Consumption of final

energy by end-use has to match with the stock of devices, their power and their duration of use. If

the stock is known with sufficient details, correlations and simple calibration procedure in

spreadsheets allow quite robust split of all energy flows by end-use sectors.


125

For example it is necessary to split the fuel consumption between end-uses (space heating, space

cooling, lighting, etc.) using a fractional share or split fuel consumption by end-use and by building

type. Figure 4-8 shows how to split the final energy consumption (the data in the example are taken

from EUROSTAT) by end-use using a fractional share. For example the residual fuel oil (RSDOIL,

cell E16) is split between Space heating (0.75, cell E22), Water heating (0.24, cell E24) and

Cooking (0.01, cell E26).

Table 4.2: Global summary energy balance for 2002 [IEA, 2004] (a)

2002 (+) in Mtoe Coal Crude

oil

Oil

Prod.

Nat.

Gas Nuc. Hydro Bio.

Elec.*,

Other TOTAL

Indigenous Production. 2403 3647 - 2169 591 577 1118 68 10573

Imports 447 2072 740 584 - - 1 45 3889

Exports -436 -1947 -813 -583 - - -2 -44 -3824

Stock Changes -12 -2 16 3 - - 0 - 6

Total Primary Energy Supply 2402 3770 -56 2173 591 577 1118 69 10643

Intl. Marine Bunkers - - -146 - - - - - -146

Transfers - -104 119 - - - - - 15

Statistical Diff. -21 -14 8 -5 - - 0 0 -31

Electricity Plants -1404 -28 -212 -447 -578 -577 -31 1172 -2105

CHP Plants -178 -1 -30 -258 -13 - -32 282 -229

Heat Plants -62 -1 -17 -86 - - -9 148 -28

Gas Works -11 - -4 8 - - - - -7

Pet. Refineries - -3642 3618 - - - 0 - -24

Coal Transformation -155 0 -3 0 - - - - -158

Liquefaction Plants -18 11 0 -8 - - - - -14

Other Transformations 0 30 -28 -4 - - -44 - -45

Own Use -45 -9 -207 -201 - - -2 -143 -606

Distribution Losses -2 -3 0 -21 - - - -145 -171

Total Final Consumption 505 11 3043 1153 - - 1000 1384 7095

Industry Sector 382 11 603 515 - - 160 572 2242

Transport Sector 5 0 1746 57 - - 8 20 1837

Other Sectors 106 0 504 580 - 832 792 2814

Non-Energy Use 12 - 190 - - - - 201

*: Other includes: solar, wind, geothermal in the primary section, electricity and heat in the use section.

+ The Total Final Consumption (TFC) is equal to the original IEA data, but not the Total Primary Energy Supply

(TPES). Here non-fossil electricity has been converted to primary equivalent by using the same average fossil efficiency

of 38.79%. When the physical energy principle is used – as the IEA statistical office does – nuclear weighs three times

more than hydro, although its electric output is nearly the same. (a) The methodology of energy balances is well

explained in chapter 7 of the “Energy Statistics Manual” issued by the statistical office of the International Energy

Agency, downloadable from http://www.iea.org/textbase/nppdf/free/2005/statistics_manual.pdf.


126

Figure 4-8 Split of fuel consumption

Once the splits are assigned, the commodity flows are quantified further to the level of useful

energy, taking into account the efficiencies of end use devices, which thereby determines the

implied stock in place, and thereby the resulting energy service demands. Analysts need ingenuity

and a lot of patience in order to complete the partial quantification made available by energy

statistical offices and ready it for the model66

.

4.3.2 Energy technology and end-use devices

Machines and devices are the most visible and durable part of any energy system. Thousands of

them extract, transport, transform, and distribute energy goods. Million of devices satisfy the

demand for energy services using energy goods. For the analyses, devices and processes similar by

type or function are usually grouped in clusters. These clusters are commonly addressed as “energy

technologies”.

Describing an energy system with ten technologies is impossible; one hundred technologies give a

very rudimentary image of the system or a detailed image of a sub-system; a real system can be

fairly well described with a number of technologies approaching one thousand or even more. Each

of these technologies is normally characterised in all aspects relevant for the analysis: technical,

economic, environmental (see an example in Table 4.3).

66

The image in the figure is a print screen taken from a VEDA template. As in several other cases, the VEDA users‟

shell helps the expert in the quantification of the system, before starting the modeling phase.


127

The technical parameters associated with technologies include efficiency, availability factor(s),

commodity consumptions per unit of activity, shares of fuels per unit activity, technical life of the

process and contribution to the peak equations. The efficiency, availability factors, and commodity

inputs and outputs of a process may be defined in various flexible ways depending on the desired

process operation, on the time-slice resolution chosen for the process and on the time-slice

resolution of the commodities involved. Certain parameters are only relevant to special processes,

such as storage processes or processes that implement trade between regions, but besides these two

special types of processes most of the parameters and features are the same and available for all

processes. A large amount of this information is used by TIMES to represent quantitatively the

system (see for instance Appendix C), all of which may be easily entered through VEDA-FE.

When data refers to existing technologies (see example in Figure 4-9), technical and economic data

are generally reliable, because they can be gathered directly from manufacturers and their

catalogues. However, it is difficult to assemble significant average values for an average technology

over a country because technologies labelled in the same way may have quite different technical

contents. It is more difficult to assemble reliable data on the stock of existing technologies and their

characteristics by year of construction / installation. This is a critical point because this information

embodies the past history of the system and conveys to the analyst of the technological content of a

system the same information that econometricians gather from time series of macro-economic data.

From the analysis point of view, the problem is transforming the energy flows in each cell of the

energy balance into the compatible stock of existing technologies. It normally happens that the

stock of a set of technologies, gas power plants, and their efficiencies, are not compatible with the

amount of natural gas allocated in the balance to electric production or to the amount of electricity

produced. It takes patience and experience to assemble a compatible set, and at times requires

choosing between the energy balance and other data sources.

Table 4.3: Example of data needed for characterising energy technologies

General Technical Economic Environment Labour & Material

Refe-rences

Technology Available size Currency GHG emissions Material: Title Technology Sector Existing Capacity Investment Solid waste Steel Author Data quality Construction time Fixed O&M Liquid waste Concrete Editor Technical availability

Technical life Variable O&M Gaseous waste ..... Type

Commercial availability

Max availability Fuel Acustic impact ..... Year

Prototype Average availability

Total ex. Fuel Land use Labour: Access

Commercialization Energy input Total incl. fuel - Construction

Market share Energy output Decommission. - Operation


128

Figure 4-9: Simplified characterisation of existing Power Plants in Tuscany, Italy

When the analysis refers to the future, the characterisation extends to technologies that could

become part of the energy system (see Figure 4-10). In this case the problem is uncertainty (about

cost, efficiencies, etc.) or ignorance (about the amount that will be built and used). However there

are reasonable estimates available from various sources67

and existing MARKAL/TIMES models.

4.3.3 Emissions

The Emission Inventory of a Country is designed to provide a comprehensive guide to the state

atmospheric emissions. Figure 4-11 shows an example of emission inventory from CORINAIR for

SO2, NOx, NMVOC, CH4, CO, CO2, N2O, and NH3 by activity. Figure 4-12 shows an example of

emission inventory in the United State for PM10, PM2.5, NOx, SO2 and NH4.

Figure 4-10: Simplified characterisation of new Power Plant in Italy

67

See for instance the important assessment provided by the International Energy Agency with the “Energy

Technologies Perspective 2006” report and the underlying global model.


129

The problem for the analyst here is to integrate the information conveyed by inventories and in

general environment statistics with the energy flows reported by the balances and the stock of

technologies. Normally the reporting categories do not match, and when they do, quantities

calculated from energy flows and unit emission coefficients do not coincide with the values

reported by environment statistics. Here too it takes time to understand what different assumptions

cause any discrepancies, and what average values makes the quantitative characterization consistent

from all points of view. Via the Excel spreadsheets (VEDA templates) the fuel, sector or technology

emissions rate can be entered, to assist with achieving consistency between energy balances, and

technology inventories.

Figure 4-11 Source: CORINAIR – Emissions by category


130

Figure 4-12 Source: U.S. EPA 1998 – Emissions by category

4.3.4 Quantifying the economic dimension of the system

Unit prices of most primary and final energy carriers are reported as current statistics. Usually

though, the detail in the statistics is much greater than for energy balances: many more fuel types

are covered and much more frequently (daily or hourly prices can be found). Figure 4-13 and Figure

4-14 show two examples of such statistics.


131

Figure 4-13: Fuel Retail Prices, 1/2 (US$/Unit) 68

How do these market prices match with the quantity of energy flowing through the system and the

stock of technologies transforming one commodity into another? Assuming the theory of economic

equilibrium as starting point for the analysis, market prices should represent the point where supply

and demand curves match.

68

Taken from the Key World Energy Statistics of the International Energy Agency.


132

Figure 4-14: Fuel Retail Prices, 2/2 (US$/Unit)

The problem is to build such curves for each commodity and sector, i.e. approximately for each cell

of an extended energy balance. Traditionally an analytical approach is used, where the parameter of

exponential curves are estimated through econometric analyses over time series or cross sections. In

the energy sector, where the technological content is very important, it is both advisable and

possible to build supply and demand curves in an alternative way. Starting from the technical and

economic data of each supply technology, it is possible to build bottom-up, stepwise (inverse)

supply curves (see Figure 4-15). Production cost result from adding the main cost components –

investment annuities, fuel and waste disposal costs, other operating and maintenance costs,

dismantling. Quantities are given by the existing stock of plants / processes.


133

Figure 4-15 Example of step-wise (inverse) supply-demand curves for electricity

Figure 4-15 shows the costs, the fixed and variable operating & maintenance costs, the waste costs

and the fuel costs associated with a suite of competing technologies looking to provide the

requested demand. The investment costs (in green) are related to the unit of capacity newly installed

by the model. For instance, for a thermal power plant the investment cost should be MEuro/MW

installed. The fixed operating & maintenance (O&M) costs are related to the unit of capacity

installed and are paid in each year for all the technical life of the technology. For example for a

thermal power plant the fixed O&M should be MEuro/MW/year. The variable operating &

maintenance (O&M) costs are related to the unit of produced activity in each year. For example for

a thermal power plant the fixed O&M should be MEuro/GWh produced or MEuro/PJ produced.

The fuel costs (in yellow) are related to the unit of consumed fuel in each year. For example for a

thermal power plant the fixed O&M should be MEuro/PJ consumed. The waste costs are related to

the dismantling of the produced wastes, from both the construction and working phases.

The similar approach is followed to build the (inverse) demand curve. Only the demand curves for

energy services approximate analytical curves and express the preferences of energy consumers.

In a perfect economic equilibrium market the price of each commodity, as reported above in Figure

4-13 and Figure 4-14, should be equal to the marginal value. Duality theory does not necessarily

indicate that the marginal value of a commodity is equal to the marginal cost of producing that

commodity. For instance, the equilibrium point shown in Figure 4-15 does not correspond to the


134

cost of any marginal supply technology69

, since it is situated at a discontinuity of the inverse supply

curve. In this case, the price is precisely determined by demand rather than by supply70

, and the

term marginal cost pricing (so often used in the context of optimizing models) is incorrect. The term

marginal value pricing is a more appropriate term to use71

.

It happens, indeed rather frequently in the energy sector, that market prices are different from

marginal values. Economists explain these differences with market imperfections, resource rents,

CO2 permit prices, energy policies (e.g. nuclear, renewable policies), etc. Some of them can be

captured by systems analyses of this type, others such as oligopoly rents or market power cannot.

On the other hand, it is important to remember that some impacts of energy use, which can be

captured by these type of systems analyses, are currently not fully reflected in market prices (e.g.

external / damage costs).

The TIMES model generator uses the input parameters of each element of the RES, including the

demand response elasticities, and builds automatically all the set of (inverse) supply and demand

cost curves. The VEDA users‟ interface and input/output spreadsheets help calculating the

production cost of each technology „ex ante‟ (i.e. based on the parameter of a single technology)

and „ex post‟ (i.e. based on model results).

4.4 The systems analysis approach: control

This section highlights the last step of energy systems analysis, its real “raison d'être”. In fact

policy-makers and decision-makers support energy systems analyses because they need to identify

effective policies and understand the effectiveness of their strategies before taking action. After

identifying and quantifying the system as outlined above, analysts carry out mental experiments in

order to understand where the system tends to go and how it reacts to different control policies.

This section first explains why it is necessary to represent the system and its development in a

model, i.e. in a set of interrelated mathematical variables representing the state of the system.

Secondly it illustrates how the system boundaries and the external “world” are represented by two

sets of exogenous variables, uncontrollable and controllable. Thirdly it lists possible targets of

decision making in the energy sector. This is followed by examples of policy instruments that can

be used to control the energy system and reach the target. Then it explains how the model and

different assumptions on the exogenous variables are used to carry out mental experiments and to

build scenarios. Some scenarios explore how assumptions of the uncontrollable variables determine

69

The marginal technology of a market is the technology that supplies the most expensive unit of commodity that has

been exchanged. In some cases, it can be linked to the cheapest unit of demand satisfied by the market. 70

It‟s daily life experience shows that prices are also influenced by demand side reactions – e.g. prices of petroleum

products coming out of the same refinery complex are priced differently depending on the demand: higher demands for

light fuels compared to heavy ones. 71

It is important to note that marginal value pricing does not imply that suppliers have zero profit. Profit is exactly

equal to the suppliers‟ surplus, and it is generally positive. Only the last few units produced may have zero profit, if, and

when, their production cost equals the equilibrium price, and even in this case zero profit is not automatic.


135

the range of possible development paths of the systems (exploratory scenarios). Other scenarios

evaluate whether the system can reach the desired targets and what set of measures are capable of

bringing the system there (policy scenarios). Eventually this section reminds the importance of

comparing the scenarios and to infer robust and hedging strategies.

A paragraph with the description of how analysts interface with policy- and decision makers should

have been added. It should illustrate what problems are presented to the analysts and how analysts

present their advice. It should explain how policy-makers and analysts together assess the likelihood

of each scenario, the trade-off between target objectives and the robustness of strategies. It should

also warn analysts that all control strategies that they elaborate are questionable, and in fact

questioned72

. But this has been omitted since the interface between policy-makers and analysts

depends too much on national and local circumstances.

4.4.1 Preparation of the mental experiments

4.4.1.1 Representation of the system in a model

Policy-makers request analysts to identify and quantify the mix of control policies that guide the

system towards the desired targets, keeping into account possible developments of the non-

controllable exogenous variables and their uncertainties. Analysts base their answers on mental

experiments with the system – actual experiments are not possible. Since the system is complex and

its future development depends on many variables, the mental experiments are carried out through

models.

System models range from non-quantitative verbal descriptions to millions of equations complexly

interrelated to be solved by powerful computers. Furthermore it is advisable to use more than one

model in order to analyse different issues of the same system and obtain the variety of results that

are needed for the planning process. Despite the complexity and ingenious relationships embedded

in a model, one should never forget that a model is always an abstract and often simplified

description of the real-world, in most cases lagging to fully capture all dimensions of the problem.

Instead of using informal mental models, analysts rely on formal mathematical models because the

policy-makers request:

- Quantitative answers;

- Reliable and established methods;

- Transparent assumptions on framework, structure and data;

- Internally consistency;

- Regular updates when new relevant information are available;

72

What TIMES proposes is the “best” outcome from a least-cost perspective. It is not infrequent that analyses and

control strategies suggested by the analysts are rejected. Often this is due to the difficulty of explaining in simple and

logical terms the reasons behind the suggestions.


136

- Reproducible mental experiments, and

- Iterations with new assumptions, towards new targets and for new policies.

Even though developing, learning and applying computer models are time-consuming efforts, it

becomes an essential step of the planning process in order to deal with the complexity and

uncertainty of energy systems development.

As in traditional engineering, here too control strategies are elaborated through quantitative

(mathematical) models of the system. However energy systems, energy models and energy policies

models are completely different from the traditional engineering systems, mathematical models and

control strategies. Energy systems are only partly governed by technical laws. A large, if not the

largest, part of the system is governed by social, economic and behavioural laws. Since the

development of energy systems results from independent and free choices of billions of “agents”, its

configuration 20 or 50 years ahead is unknown. Since the laws determining the development of the

system are not deterministic, deterministic models represent the system development in an

incomplete way. The images of the future that are calculated through these models are “what-if”

projections, but neither forecasts nor predictions.

Furthermore, contrary to traditional engineering control models, energy systems development

models cannot be validated by means of experiments. The validation of the model in the base year73

(for which statistical values are available) ensures continuity between past events and future

assumptions but cannot ensure that the model represents correctly the system twenty or fifty years

ahead.

The model has to represent the system according to the laws that govern it. The state (or decision)

variables have to encompass all the elements identified and quantified in the previous steps. The

attached box illustrates how MARKAL-TIMES models help achieving these objectives.

Furthermore in the model the system must take different development paths and react to possible

different developments of the exogenous variables.

73

Describing the underlying energy system by necessity starts with the calibration procedure which is partly embodied

in the VEDA procedures. It requires tuning, and fine tuning, all exogenous variables in the base year till all state

variables of the model – quantity and prices – match the statistical data. The calibration is somewhat analogous to the

econometric estimate on available time series of the parameter of traditional economic models: it embodies in the

technical economic model the information about the past in the much more robust form of stock of technologies. It has

to be repeated whenever new energy balances and other important statistical data are made available, or when moving

the base (1st) year of the model. A precise calibration to the base year is not mandatory to understand the importance of

uncontrollable variable or the effectiveness of control policies, but it largely increase the credibility of models and

analyses for non experts.


137

4.4.1.2 Uncontrollable and controllable exogenous variables

Endogenous variables represent elements of the system and describe its state; exogenous variables

represent elements not included in the system. The content of the two categories depends on the

definition of the system boundaries. Exogenous variables of a system may become endogenous in

an extended system. In order to carry out the mental experiments of energy system control, the

schematization of what is not included in the system is an important complement to the model itself.

The main problem is that several exogenous variables, such as the amount of ultimate fossil

resources in place, the availability of renewable resources, the efficiency of a thermal cycle or the

Box: MARKAL-TIMES technical economic models of energy technology systems

Using TIMES-VEDA analysts represent their energy systems with 4E (energy, economic, environment,

engineering) technical-economic equilibrium models. According to the system and the target, TIMES

models build least-cost, or partial equilibrium, or general equilibrium development paths, with perfect

foresight or myopic (details are given in the first chapters of the Users‟ guide, downloadable from

http://www.etsap.org/documentation.asp).

The TIMES energy economy is made up of producers and consumers of commodities such as energy

carriers, materials, energy services, and emissions. TIMES models, like most equilibrium models, assume

competitive markets for all commodities. The result is a supply-demand equilibrium that maximizes the

net total surplus (i.e. the sum of producers‟ and consumers‟ surpluses). TIMES may, however, depart from

perfectly competitive market assumptions by the introduction of user-defined explicit constraints, such as

limits to technological penetration, constraints on emissions, exogenous oil price, etc. Market

imperfections can also be introduced in the form of taxes, subsidies and hurdle rates.

Using these as inputs, the TIMES model aims to supply energy services at minimum global cost (more

accurately at minimum loss of surplus) by simultaneously making equipment investment, operating them,

supplying primary energy, and energy trade decisions, by region. For example, if there is an increase in

residential lighting energy service relative to the reference scenario (perhaps due to a decline in the cost of

residential lighting, or due to a different assumption on GDP or population growth), either existing

generation equipment must be used more intensively or new – possibly more efficient – equipment must

be installed. The choice by the model of the generation equipment (type and fuel) is based on the analysis

of the characteristics of alternative generation technologies, on the economics of the energy supply, and on

environmental criteria. TIMES is thus a vertically integrated model of the entire extended energy system.

In TIMES – like in its MARKAL forebear – the quantities and prices of the various commodities are in

equilibrium, i.e. their prices and quantities in each time period are such that the suppliers produce exactly

the quantities demanded by the consumers. This equilibrium has the property that the total surplus is

maximized. The fact that the price of a commodity is equal to its marginal value is an important feature of

competitive markets.

A mature TIMES model may include several thousand technologies in all sectors of the energy system

(energy procurement, conversion, processing, transmission, and end-uses) in each region. Thus TIMES is

not only technology explicit, it is technology rich as well. Furthermore, the number of technologies and

their relative topology may be changed at will, purely via data input specification, without the analyst ever

having to modify the model‟s equations. The model is thus to a large extent data driven.

It should also be noted that both MARKAL and TIMES are “open,” in that the data, transformations

thereof, mathematics and model generator source code are all fully transparent and documented.


138

contribution of different GHG to radiative forcing, exert high influence on the behaviour of the

system but are not, or negligibly, influenced by policies and measures. Other exogenous variables,

such as the discount rate, the prices of energy goods, the efficiency of the devices available on the

market, or emission standards, strongly depend on policies and measures. For the purpose of

controlling the system it is useful to distinguish the exogenous variables in uncontrollable and

controllable. The level of controllability of the energy system depends on the number and

importance of the state variables that are influenced directly or indirectly by the control exogenous

variables.

The most important set of exogenous assumptions for economic growth models in general, and

technical economic models in particular, are the characterization of technological change74

.

Different assumptions on technical and economic progresses of commercially existing and new

technologies75

determine quite different development paths of energy systems. The technological

innovation in the energy sector is only partly controllable through R&D policies (learning by

searching), while the deployment of new and better technologies is more affected by long term

policies on information and regulation, tax and subsidies.

A related set of exogenous assumptions is the amount of energy reserves and ultimate resources.

Here too something that should be uncontrollable by definition, such as the ultimate oil resources,

produces variables, such as the amount of recovered oil, that are partly dependent on system prices

– because the reservoir cultivation techniques, the technology to displace the oil in place, and the

number of perforation depend on the available investments.

Another set of exogenous assumptions is the future development of the demand for energy, be it

primary, final, useful or energy service. Several studies on historical time series or cross section

analyses of macro-economic indicators help making demand projections based upon selected

“drivers” such as population, households, GDP, etc76

. Demands are projected through simple

formulas such as77

:

topricesElasticitytodriverElasticity pricesDriverDemand *

74

In the traditional general equilibrium models, technologies as a group are represented by production functions, In

technical economic models, each technological group is explicitly characterised by parameters such as year of first

availability (e.g. hydrogen technologies), energy efficiency, capacity factors, costs, emissions, etc. 75

In the literature the phenomenon of technical improvement, on which actually all economic growths depend, is called

„technology learning‟. In this context it is not necessary to distinguish among different types of learning (due to R&D or

market forces) or alternative ways of representing the phenomenon in a model (exogenous of endogenous, ETL): it all

ends up in different assumption on characterization parameters, some additional equations and more complexity. 76

For example, the TIMES models constructed for the NEEDS (New Energy Externalities for Developments in

Externalities) and EFDA (European Fusion Development Agreement) used the GEM-E3 (General equilibrium Model

for Economy, Energy and Environment) general equilibrium model to generate a set of coherent (total and sectoral)

GDP growth rates in the various regions. Note that GEM-E3 itself uses other drivers as inputs in order to derive GDP

trajectories. These GEM-E3 drivers consist of measures of technological progress, population, degree of market

competitiveness, and a few other perhaps qualitative assumptions. 77

The own price dependence is embodied in the TIMES model, once the elasticity is declared.


139

Although the behaviour of the independent drivers is hardly controllable through policies, the

dependence on prices via elasticities makes demands partly controllable through prices.

Projecting the price of import / export commodities should be linked to the assumed availability of

each resource. But rules are not easy to defend with an international market that has shown a ten-

fold increase of crude oil price, from 10 to 100 $/bbl in six years. Furthermore exogenous price

projections may include taxes and subsidies, a policy dependent component.

4.4.1.3 Objectives and targets

Controlling the future development of energy systems is dictated by several needs. In order to

ensure people that they will continue to be able to satisfy their demand for energy services

(availability) governments of countries without sufficient energy resources are concerned about

security of supply from exporting countries. The twin goal is to ensure that the country does not

consume too many economic resources in the purchase of energy resources (affordability), in order

to have enough economic resources to ensure the economic development. The alternative goal is to

ensure that supplying energy services to the consumers does deprive them of the right to enjoy of

public goods such as health and environment (sustainability).

The problem starts when these generic goals have to be converted into precise targets. For instance

some country may want to develop its own domestic resource in order to supply at least 55% of the

domestic demand in 20 years. Another region has the target to reduce acid deposition, hence

emission of precursors such as sulphur and nitrogen oxides, to 20% of the present values in 30 years

in order to preserve domestic forests. The global community could establish a target to half carbon

dioxide emissions in 50 years. Are these targets feasible? At what cost? How?

In establishing the targets, the problem is that sometime they are contrasting. Although in general

they might be shared by all policy-makers, the emphasis and the numerical targets are different

depending on who formulates the goals: energy, environment or economics policy-makers. When

negotiations start, between different policy-makers in a country or different countries, with the aim

of agreeing on some common target, it is necessary to know in advance what are the trade-offs

between different objectives to be reached78

.

4.4.1.4 Policies instruments and specific measures as control variables

The development of energy systems can be controlled by several types of policy instruments.

Traditionally domestic policies and measures can be grouped as follows:

78

The methodologies offered by ETSAP support these sort of analyses and provide to policy-makers with evaluations

whether a target is feasible, how much it costs in terms of the other objectives, and, as explained below, what technical

economic paths required to achieve each target.


140

Legal Instruments: command and control; codes to improve the thermal insulation of

buildings, minimum efficiency standards for end use devices including cars, integrated

prevention and pollution control over existing and new plants, traffic restrictions, etc.;

Direct Investments: energy research development and deployment (RD&D), procurement…

Other Economic Instruments: tax and subsidies, liberalisation, unbundling, etc.;

Voluntary Agreements: rational use of energy industry/municipalities, energy audits in

industry, services, buildings, phase out of the less efficient end use devices, etc.;

Diffusion of Information: monitoring mechanism, energy labels of end use devices, and

Examples of measures in the field of energy R&D and technology development policy are:

- Analysing competitiveness of technologies or energy chains (e.g., gas or district heating

grid) under different economic assumptions and market barrier removal;

- Assessing bundles of competing and/or of complementary technologies, rather than stand-

alone evaluations;

- Evaluating the impact of technological progress and aids to research, development, and

deployment (learning-by-doing, learning-by-researching);

- Calculating supply and demand (cost) curves; and

- Adding a dynamic dimension to Life-cycle Analysis (LCA, cradle-to-grave).

Examples of measures in the field of energy policy are:

- Mandatory micro-measures in each sector: building code, building retrofit programs, modal-

split incentives in freight and passenger transports, energy efficiency programs, etc. vehicle

standards;

- Energy taxes, investment subsidies (green and white certificates; clean / efficient

technologies);

- Energy security evaluation (e.g., measured by oil/ gas/ nuclear fuel imports energy options

evaluation);

- Fuel and technology mixes in general;

- Education, information, and

- Social constraints: e.g. nuclear.

Examples of measures in the field of environment emissions and climate change mitigation decision

making are:

- Emission taxes, incentives to non polluters; tax redistribution issues;

- Emission Cap-and-trade systems: global or partial coverage, multiple bubbles, etc.;

- Hybrid system: caps + ceiling on emission price;

- Emission intensity standards and regulations;

- Internalising environmental externalities;

- Alternative allocations of emission rights to regions, sectors; lumped allocations versus

output-based allocations, and

- Energy intensive materials & urban solid waste management.


141

But transforming any of the above types of policy instruments into a particular measure is not easy

because of the long term implications far beyond the specific target pursued79

. And in all these

circumstances there are hard questions to be addressed. What is a reasonable balance domestic

mitigation and purchase of emission permits? What is the cost of complying with national/regional

targets? What is the cost of reducing acid deposition emissions to comply with your regional

targets?

4.4.2 Carrying out the mental experiments: scenarios80

As mentioned before, scenarios explore different development paths of the system in the space of

possible (future) events. They all share the same starting point, the actual energy system as

quantified by existing statistical information and their integration. They assume different

development of either the uncontrollable variable (alternative scenarios) or the controllable ones

(policy scenarios) or both.

The mental experiments traditionally follow this sequence of steps:

- Observation of the preferred development path of the system without exerting any

(additional) control (base case or business as usual scenario – BAU)81

;

- Exploration of alternative development trajectories in the space of events through different

assumption of the non-controllable exogenous variables (alternative exploratory scenarios);

they are intended to explore the effect of uncertainties on exogenous variables; exploratory

scenarios are a sort of a sensitivity analysis on assumptions and identification of key- factors

that might affect the future of the energy economy;

- Experimentation of different control strategies and policies capable of bending the each

uncontrolled development paths (exploratory scenario) towards the desired target(s) (policy

scenarios); and

- Selecting the control strategies that most frequently (if not for all scenarios) make the

system reach the target (robust strategies) and finding out hedging strategies82

.

79

ETSAP has developed tools for evaluating the impact of several energy related decisions in the four main

dimensions: energy, engineering, economic and environment. On the contrary, ETSAP tools and in general analyses of

energy system offer little help to understand: demographic dynamics, economic development and GDP growth (unless

TIMES-MACRO is employed); how many degrees will temperature increase due to climate change; the demand for

energy services (unless TIMES-ED or MACRO are employed); the mining cost in 30 years; emissions technology

factors, or engineering technology data. 80

The literature on energy scenarios is very ample. Some base elements on scenarios, the related bibliography, and the

description of some relevant long term global scenarios are reported by G.C. Tosato, “Global long-term energy

scenarios: lessons learnt”, IPP Report No. 16/13, March 2007. The report is downloadable from:

www.ipp.mpg.de/ippcms/de/kontakt/bibliothek/ipp_reports/IPP_16_13.pdf. 81

This scenario normally shows quantities and prices in the base year different from the statistical values. The distance

of the actual system from the economic equilibrium indicates the weight of non technical economic laws and factors in

the system. Part of the gap can also be viewd as the effect of different „market failures‟and sometimes suggests where

distortion can be removed. 82

VEDA Back End is very useful in extracting relevant information and presenting them synoptically (in a footnote to

several scenarios).

http://www.ipp.mpg.de/ippcms/de/kontakt/bibliothek/ipp_reports/IPP_16_13.pdf


142

Since the range of uncertainty, or arbitrariness, about data characterising the future is very high83

,

the effect of uncertainty is explored through different assumptions only on the most important set of

exogenous assumptions. The modellers‟ challenge is to explore the widest possible areas of

uncertainties with a limited number of scenarios.

In general, to hedge against uncertainties of far-reaching projections or otherwise doubtful input

data, different paths of future development involving more complex changes to the model database

may be analysed, i.e. variations of more than one model parameter value are necessary. For

example, it will usually not suffice to vary the price of coal at one point in time - rather the period

values throughout the time frame will have to be changed for an assessment of a “price scenario.”

Furthermore, taking into account the interdependency of prices of the various energy carriers, an

alternative price path for coal should trigger (different) price changes for other energy carriers as

well. Therefore, an event tree is often used to examine which possible combinations of exogenous

variables projections are feasible, meaningful of interesting (see Figure 4-16).

Consistent hypotheses of the prices of relevant energy carriers over the study time frame have to be

assessed in separate model runs and the results analysed thereafter. Each hypothesis must be based

upon predicted global economic developments, as well as conditions specifically connected to the

area of investigation. This is perhaps the most difficult step in scenario analyses.

Figure 4-16: Selection of scenarios in an events tree: an example

83

In fact even for presently commercially available technologies it is difficult to reduce what social scientist call

“technology controversies”.


143

4.4.3 Robust and hedging strategies84

In this area, a strategy can be defined as a set of policy capable of bringing the system towards the

intended target. After building policy scenarios for each exploratory scenario, by comparison across

scenarios, analysts identify the set of policies capable of bending each non intervention scenario

toward the target. Some policies appear once or twice, others are effective in all (or nearly all)

exploratory scenario. In this sense, a “robust strategy” is the set or policies capable of bending the

future towards the intended scenario in the widest range of evolution of uncontrollable variables.

Analysts face a more complex problem when a parameter is uncertain in the near future and will be

known later. A typical example is given by the “climate sensitivity” (Cs)85

. Its value is now

uncertain: 1.5 ºC, or 3ºC, or 5 ºC, or 8ºC according to the expert; the most we know now is the

probability of each value (see Figure 4-17) and that by 2050 the value of Cs will be known

precisely.

The traditional way of exploring future possibilities would be to compile four separate scenarios,

each one with a different value of Cs. In this way we obtain four different trajectories (see Figure

4-18) and strategies. But this approach leaves policy makers with the problem of deciding which the

preferable strategy is in the short term. In these situations, policy makers have to act now and

cannot wait until uncertainty is resolved before acting.

Figure 4-17: The event tree of possible climate sensitivity values

84

This paragraph reproduces the content of a presentation given by Richard Loulou at the ETSAP Workshop held in

Brasilia on November 19-23, 2007. The graphs are taken from the results obtained by Richard Loulou, Amit Kanudia,

Maryse Labriet, Uwe Remme in the frame of the ETSAP participation to the Energy Modelling Forum 22: “Climate

Policy Scenarios for Stabilization and in Transition”, Working Group on Hedging. 85

The climate sensitivity is the temperature increase corresponding to a doubling of pre-industrial CO2 concentration.


144

Figure 4-18: Example of exploratory scenarios without a hedging strategy

A better strategy is unique in the short term, but acknowledges future uncertainties: such strategy is

called a “hedging strategy”. A hedging strategy is obtained if the policy maker includes in his (her)

thinking all possible future scenarios to produce a single strategy, until the uncertainty is resolved.

This is normal practice in several fields: a homeowner hedges against many possible futures (fire,

theft, hurricane) by insuring his house; a businessman hedges against exchange rate fluctuations by

buying a future for a currency. Such strategy is illustrated in Figure 4-19. The dotted lines represent

emissions for a hedging strategy: it has a single trajectory before 2050, and multiple trajectories

after uncertainty is resolved.

Figure 4-19: Example of hedging strategy


145

5 References

R. Loulou, U. Remne, A. Kanudia, A. Lehtila & G. Goldstein, Documentation for the TIMES

model - Part I, II, III, ETSAP, 2005.

R. Loulou, A. Lehtila, Stochastics TIMES, ETSAP, 2005.

R. Loulou, A. Lehtila, TIMES Damage function, ETSAP, 2005.

U. Remme, M. Blesl, Documentation of the TIMES-MACRO model, ETSAP, 2006.

IEA – Annex 33, Advanced Local Energy Planning Guidebook (ALEP).

IEA (2006), Key World Energy Statistics 2006.

OECD/IEA (2003), Eenergy to 2050 - Scenarios for a Sustainable Future

OECD/IEA (2006), Energy Technology Perspectives – Scenario & Strategies to 2050.

Nebojsa Nakicenovic (IIASA), Energy Primer.

European Communities (2006), Key figures on Europe – Statistical Pocketbook 2006.

U. Remme, Basic elements and equations of TIMES, ETSAP Workshop, 2006, Stuttgart.

S. Kempe, Modelling of CHP plants in TIMES, TIMES training, 2006, Stuttgart.

G.C. Tosato, “Global long-term energy scenarios: lessons learnt”, IPP Report No. 16/13, March

2007

getting started with times-veda version 2 - … · getting started with times-veda version 2.7 may...

Documents