TIMES Reference Energy System

PriceA

Demandcurve

SupplyCurve

Equilibrium

PE

1

Quantity QE

Figure 3.1Supply demand equilibrium

3.3 TIMES Reference Energy System

To generate a system model, TIMES needs various data inputs from the user pertaining to technology, commodity, and commodity flow. Broad inputs and outputs for a TIMES based model are outlined in the Figure 3.2. Commodities can be various energy carriers (electricity, heat,etc.), materials, monetary flows, and emissions. Technologies are the representation of various physical processes which either produce a commodity (mining, import, etc.) or transform some commodity into others (power plants, vehicles, demand devices, etc.).

Finally, commodity flow is the amount of commodity consumed or produced by a process; in other words they are the links between process and commodities. The input data to TIMES can be either qualitative (list of various types of commodities and processes) or quantitative (techno-economic parameters).

Technical parameters associated with processes are efficiency, availability factor, technical life, construction lead time, commodity consumption per unit of activity, etc. Economic parameters include various costs related to investment, operation and maintenance, and dismantling. Other than these parameters, taxes/ subsidies and various bounds related to investment, capacity, and activity of a process for a period/ region can be defined. Technical parameters for commodities can be overall efficiency/ loss, and declaration of traceable time slice. For final demand commodities, additional annual demand projection is needed over

Energy conversion

Transmission distributionand

Demand technologies Current Resource

Stock, Future Demand, Technology Description, Policy

Energy generation, capacity, fuel mix, investment,

emission Energy System Model

Input Output

Figure 3.2Input-outputs of a TIMES based energy system model

the model horizon. If the demand commodity is associated with intra-annual time slices, user need to provide the corresponding demand curve pattern also. Economic parameters associated with a commodity flow are additional costs, taxes, and subsidies on the production of a commodity. Technical parameters are share of a particular commodity within a input/

output commodity flow group, efficiency, emission rate by fuel,etc.

The inter relationship between the commodity, processes, and commodity flow creates a directed graph which is termed as RES (reference energy system). In a TIMES RES, commodity flows are links between process and commodity. Though a RES can portray the whole picture of an interconnected energy system starting from resource extraction/

procurement to end use, it can also exclusively focus on a specific sub sector. In general, RES diagrams can be helpful to track the material or commodity flow between various processes.

3.3.1 TIMES Optimization Framework

TIMES based models minimize the discounted sum of the annual costs (investment, operation

& maintenance, energy import, export, and delivery, resource extraction, tax, and subsidies, etc.) while satisfying several constraints over the modeling horizon. To calculate the net present value of the system cost, the model first calculates regional yearly total system costs over the modeling horizon. Further, the costs are discounted to the base/ reference year for every region (Equation 3.1).

NPV =

R r=1

∑ ∑

y∈years

(1+d_r,y)^{REFY R−y}∗ANNCOST(r,y) (3.1) where,

3.3 TIMES Reference Energy System 47 NPV is the net present value of the total cost for all regions (the TIMES objective function) ANNCOST_r,y is the total annual cost in region r for year y

d_r,y is the general discount rate REFYR is the reference year

R is the set of regions

YEARS is the set of cost incurring years (including past investment and dismantling costs, Salvage Value)

TIMES follows linear programming principle. By minimizing the objective function outlined above, it calculates the optimum values of several decision variables, such as new capacity addition, total installed capacity, activity level of technology, quantity of commodity consumed/ produced by a process, price of commodities,etc. Minimization of the objective function and calculation of decision variables values is subjected to satisfying several physical and logical constraints. Details of TIMES modeling platform with descriptions of each set, parameters, variables, and equations are available in the documentation [177]. Some of the key model constraints are described bellow.

• Commodity balance: In each time period, for every region, total commodity production plus imports must balance region’s consumption plus export.

• Use of capacity: Activity of a technology per time slice/ annually may not exceed its available capacity, as specified by the availability factor.

• Commodity constraints: Commodity production/ extraction limits such as emission cap or annual fossil fuels extraction bounds

• Growth constraints: Yearly growth rate of process capacity within certain bounds to avoid excessive abrupt investment in new capacity

• User constraints: User defined constraints involving any TIMES variable. Specially suitable to model policy scenarios such as, nuclear capacity phaseout, renewable portfolio standards,etc.

There has been some recent development in TIMES modeling platform to enable modelers to incorporate unit commitment and dispatch related constraints of power plants and DC power flow for electricity trading [121, 178]. Consideration of these constraints and features

may lead to more accurate dispatch decisions of generators and regional power exchange.

But, effectiveness of various constraints in planning model and their impact on computational complexity are subjected to scrutiny [16, 179]. There is need to identify specific constraints to incorporate in the planning model based on trade-off between additional computational complexity and result accuracy [19, 180]. In the present study those constraints have not been considered in the planning model. But, detailed unit commitment and dispatch constraints are considered in a separate operational model with which results of the planning model is compared.

TIMES based models are suitable to analyze long-term system development via scenarios or ‘what-if’ analysis. Scenarios are different from forecasts as they follow logical story lines involving various assumptions of future trajectories of several drivers. In TIMES, a scenario consists of various data inputs pertaining to demand and supply curves, policy definition, and descriptions of technologies. The base, or Business As Usual (BAU) scenario corresponds to the data provided for the current system development trajectory. After optimization, the model calculates the values of decision variables of BAU as well as other cases specified by the user. The decision regarding feasibility and suitability of future scenarios are taken by comparing them with the base case results.

Model Output

VEDA-FE

VEDA-^BE

TIMES

GAMS + Solvers Data and

Assumptions

Data Handling Model Generator

Result Handling Model Solution +

LMA

Figure 3.3TIMES model work-flow

3.3.2 Working With TIMES

The source code of TIMES model generator is written in GAMS (General Algebraic Modeling Language), a high level modeling language to formulate large-scale optimization problems.

As high volume of data and assumptions are involved for the model building, separate tools/

software are developed which can work as a interface between user and GAMS. Therefore,