How Mathematical Optimization Powers Decision-Making in Power Markets

Electricity markets operate across multiple decision layers—spanning long-term planning, commercial procurement, and real-time operations. While these decisions may appear operationally different, they are unified by a common foundation: mathematical optimization under technical, regulatory, and commercial constraints.

In this blog, I walk through how different optimization models are applied to solve distinct decision problems in power systems—ranging from capacity expansion and contract procurement to resource scheduling.

1. Capacity Expansion Optimization: Strategic Planning

The first and most fundamental question in power systems is:

What capacity (and which technology) should be added to reliably satisfy future demand at the minimum cost?

To address this, a capacity expansion optimization model is developed spanning a multi-year horizon (e.g., 9 years) using Mixed-Integer Linear Programming (MILP).

Core Features

• Addition of new Renewable Energy (RE) capacity, Thermal capacity, and Battery Energy Storage Systems (BESS)
• Minimize long-term portfolio tariff / system cost
• Ensure at least 95% of forecasted demand is met while maintaining operational feasibility and regulatory constraints

Outcome

This model identifies the optimal capacity mix, balancing reliability, cost, and policy constraints over the long term.

2. Optimal Contract Suggestion: Procurement Decision Support

Even with sufficient installed capacity, utilities and generators often face short-term demand deficits due to variability in demand and availability.

Given a deficit and a set of available contracts, which contracts should be procured from, and in what quantity?

Optimization Focus

• Select cost-optimal profiles and procurement quantities across bilateral, market-linked, and profile-based contracts
• Incorporate technical availability, contractual limits, and operational constraints
• Objective: Minimize procurement cost while covering demand deficit

Outcome

The model recommends cost-optimal contract profiles and quantities, supporting improved real-time coordination (RTC) and peak demand management.

3. Resource Scheduling Optimization: DA, ID & RSD

Day-Ahead & Intra-Day Optimization

Day-Ahead optimization focuses on cost-efficient base schedules, while Intra-Day optimization manages deviations in forecasts and availability.

We developed Linear Programming (LP)-based scheduling models to:

• Optimize generation schedules for Day-Ahead and Intra-Day markets
• Adapt dispatch based on technical constraints, real-time demand variation, and cleared bids

These models replace static rule-based scheduling with adaptive, optimization-driven decision-making.

Reserve Shut Down (RSD)

To evaluate the economic impact of RSD or zero-schedule decisions, an MILP-based RSD model was developed to:

• Simulate scenarios with and without RSD
• Optimize unit commitment
• Minimize OPEX while capturing fixed-cost savings

This enables decision-makers to assess whether shutting down reserve units is financially justified.

4. Cost–Benefit Analysis Across Markets

Cost–benefit analysis was conducted across DAM, RTM, and GDAM to evaluate:

• Financial impact of bidding and operating strategies
• Pre-facto vs post-facto decision outcomes
• Trade-offs between cost minimization and revenue-risk balance

This layer connects optimization outputs to realized financial performance.

5. A Unified View: Different Problems, One Optimization Mindset

Although each layer solves a different problem, they form a coherent decision flow:

Capacity Expansion
        ↓
Optimal Profile Suggestion
        ↓
Resource Scheduling (DA / ID)

Each stage:

• Has a distinct objective
• Uses mathematical optimization techniques
• Feeds logically into the next decision layer

Final Thoughts

Power markets are governed not by a single optimization problem, but by a stack of interconnected decision models—strategic, commercial, operational, and financial.

Understanding which problem you are solving is as important as how you solve it.

By treating capacity planning, contract procurement, and scheduling as separate yet connected optimization problems, utilities and market participants can achieve:

• Lower costs
• Better regulatory compliance
• More resilient operations

Optimization, when applied with the right framing, becomes not just a mathematical tool—but a decision-making engine for modern power systems.