⚡ electrical engineering

Multiobjective optimization-based design and dispatch of islanded, hybrid microgrids for remote, off-grid communities in sub-Saharan Africa

This paper presents a multiobjective optimization framework for designing and dispatching islanded hybrid microgrids in sub-Saharan Africa, demonstrating that a solar-wind-battery-diesel system optimized via particle swarm optimization effectively balances cost, emissions, reliability, and efficiency while revealing critical tradeoffs between economic and environmental objectives.

Original authors: Vineet Jagadeesan Nair

Published 2026-04-09

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Vineet Jagadeesan Nair

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are the mayor of a tiny, remote village in the middle of the African savannah. The nearest power line is hundreds of miles away, and the cost to run a wire to your village would be so high it would bankrupt the community. You need electricity for lights, water pumps, and schools, but you can't rely on the big national grid.

So, you decide to build your own mini-power plant right there in the village. But here's the tricky part: you want it to be cheap, clean (no smoke), reliable (no blackouts), and efficient (no wasted energy).

This paper is like a master blueprint for building that perfect mini-power plant. The author, Vineet, used a super-smart computer program to figure out exactly how big your solar panels, wind turbines, and batteries should be, and how to run them every hour of every day.

Here is the story of what they found, explained simply:

1. The "Recipe" for the Perfect Power Plant

The author tested different combinations of ingredients to see what makes the best "power soup."

The Ingredients: Solar panels (sun), wind turbines (wind), batteries (storage), and a diesel generator (the backup).
The Winner: The best mix turned out to be Solar + Wind + Lithium-ion Batteries + a small Diesel backup.
The Analogy: Think of it like a diet. You want to eat mostly healthy food (sun and wind) to stay fit and save money. But you keep a bag of emergency snacks (diesel) in your pocket just in case you get really hungry during a storm. The study found that using a mix of sun and wind is cheaper and cleaner than relying on just one or the other.

2. The "Brain" of the Operation (Optimization)

Building the plant is one thing; running it is another. You can't just turn everything on at once. You need a brain to decide: "Should I charge the battery now? Should I turn on the diesel engine? Or should I let the sun do the work?"

The author tested different "brains" (math algorithms) to see which one could make the best decisions.

The Contenders: They tried Genetic Algorithms (like evolving solutions over time), Simulated Annealing (like cooling metal to find the perfect shape), and Particle Swarm Optimization (PSO).
The Winner: Particle Swarm Optimization won.
The Analogy: Imagine a flock of birds looking for the best patch of berries. Each bird flies around, checks the berries, and tells the others, "Hey, I found a good spot!" The flock moves together, learning from each other, until they all converge on the absolute best spot. PSO works the same way: it sends out many "virtual birds" to search for the perfect size for your solar panels and batteries, and they quickly find the best solution together.

3. The "Goldilocks" Problem (Sizing)

One of the biggest mistakes villages make is oversizing. They think, "If we buy double the solar panels, we'll never run out of power!" But this is like buying a giant truck just to drive to the corner store. It costs too much to buy and maintain.

The Finding: The study found that you don't need to overbuild. By using the "smart brain" (PSO) to calculate the exact size needed, you can cut costs significantly.
The Trade-off: There is a balance. If you want 100% reliability (never a blackout), you need a bigger backup generator. If you want to save money, you might accept a tiny, rare chance of a blackout. The study maps out these trade-offs so the village leaders can choose what matters most to them.

4. The "Weather Rollercoaster" (Uncertainty)

What happens if it's cloudy for three days? Or if the wind stops blowing?

The Risk: The study found that the system is very sensitive to wind. If the wind stops, the village has to rely heavily on the diesel generator, which gets expensive and dirty.
The Solution: The "smart brain" can predict these dips. It charges the batteries when the sun is shining and saves the diesel for the really bad days.
The Human Factor: The study also suggests that if the villagers can shift their habits (like washing clothes or pumping water during the day instead of at night), the system works even better. It's like everyone agreeing to take a nap when the sun is brightest so the power plant doesn't have to work so hard.

5. Why This Matters for Africa

The paper focuses on Sub-Saharan Africa, where millions of people live without electricity.

The Big Picture: Extending the main power grid to these remote villages is like trying to build a bridge across the ocean just to get to a small island. It's too expensive.
The Verdict: Building these independent "mini-grids" is actually the cheapest and fastest way to bring power to these communities. It's like giving every village its own independent power station instead of waiting for a giant wire to arrive.

Summary

This paper is a guidebook for building a smart, self-sufficient power station for remote villages. It proves that by mixing sun and wind with a little bit of battery storage and a backup generator, and by using a "flock of birds" computer algorithm to manage it all, we can provide clean, affordable electricity to people who have never had it before.

It's not just about math; it's about lighting up homes, powering schools, and helping communities thrive without breaking the bank or choking the air with smoke.

1. Problem Statement and Background

The paper addresses the critical challenge of providing reliable, affordable, and sustainable electricity to remote, off-grid communities in Sub-Saharan Africa (SSA), where over 600 million people lack access. Traditional grid extension is often economically unfeasible due to high transmission costs (up to $5.5 million/MW/km) and long lead times.

The study focuses on the optimal design, sizing, and operational dispatch of islanded hybrid microgrids. These systems must balance four competing objectives:

Economic Affordability: Minimizing the Levelized Cost of Energy (LCOE).
Environmental Sustainability: Minimizing greenhouse gas and pollutant emissions.
Reliability: Minimizing the Deficiency of Power Supply Probability (DPSP) or lost load.
Efficiency: Minimizing wasted (dumped) energy and maximizing renewable energy penetration.

The specific context is a rural community in Timbila, Kenya, characterized by intermittent renewable resources (solar and wind) and a lack of connection to a macrogrid.

2. Methodology

A. System Architecture

The proposed microgrid is a hybrid system comprising:

Renewable Energy Sources (RES): Solar Photovoltaic (PV) arrays and Wind Turbines (WT).
Energy Storage (ES): Lithium-ion (Li-ion) and Lead-acid (LA) batteries.
Backup Generation: Diesel Engines (DE) and Micro-turbines (MT) running on fossil fuels.
Power Electronics: Charge controllers, inverters/rectifiers, and power conditioners.

B. Optimization Framework

The study employs a multiobjective, multiperiod global optimization framework with two distinct stages:

Design and Sizing Optimization:
- Horizon: 1 year (8,760 hours).
- Decision Variables: Number of PV modules ( $n_s$ ), number of wind turbines ( $n_w$ ), and initial rated battery capacity ( $E_{b,init}$ ).
- Objective: Minimize a weighted sum of LCOE, emissions, DPSP, dumped energy, and (1 - Renewable Energy Fraction).
Day-Ahead Dispatch Optimization:
- Horizon: 24 hours (hourly resolution).
- Decision Variables: Power output of dispatchable units (Diesel Generator $P_{DG}$ and Battery Storage $P_{BS}$ ).
- Constraint: Renewable output is non-dispatchable and determined by weather; the optimizer schedules the backup and storage to meet load.

C. Algorithm Selection

The problem is formulated as a non-convex, non-smooth, mixed-integer nonlinear programming (MINLP) problem due to discrete variables (component counts) and non-linear cost functions. The author evaluated several global optimization solvers from MATLAB's Global Optimization Toolbox:

Genetic Algorithm (GA)
Particle Swarm Optimization (PSO)
Simulated Annealing (SA)
Surrogate Optimization (SO)
Pattern Search (PS)

Result: Particle Swarm Optimization (PSO) was identified as the superior solver, offering the best balance between computational speed (63 seconds vs. >800 seconds for others) and solution optimality (lowest objective function value).

D. Dispatch Strategy

A load-following strategy was encoded for the sizing phase, prioritizing:

RES generation to meet load.
Excess RES to charge batteries.
Excess beyond battery limits to dump loads.
Battery discharge when RES is insufficient.
Backup DG activation when batteries are depleted.
Note: The day-ahead dispatch optimization revealed that a dynamic, non-priority-based strategy (simultaneous DG and battery usage) often yields better economic results than the rigid load-following rule.

3. Key Contributions

Algorithmic Benchmarking: The study provides a rigorous comparison of global optimization algorithms for microgrid sizing, demonstrating that bio-inspired methods (PSO, GA) outperform traditional gradient-based or exhaustive search methods for this specific non-convex problem class.
Regional Specificity: Unlike many studies focusing on grid-connected systems or generic locations, this work specifically targets Sub-Saharan Africa, using local climate data (Kenya) and addressing the unique economic constraints of remote villages.
Integrated Design-Dispatch: The paper moves beyond sequential sizing and dispatch by analyzing how optimal sizing changes when coupled with advanced day-ahead dispatch strategies, revealing that rigid sizing rules can lead to suboptimal operations.
Pareto Frontier Analysis: The study maps the trade-offs between economic, environmental, and reliability objectives, showing that cost-only optimization often leads to poor reliability or high emissions.
Break-Even Distance Analysis: It quantifies the economic viability of microgrids versus grid extension, calculating a break-even distance of 0.855 km for the study site, confirming microgrids are the superior choice for locations >1 km from the grid.

4. Key Results

A. Optimal Technology Mix

Best Configuration: A hybrid system using Solar PV, Wind, Lithium-ion batteries, and a Diesel Engine backup consistently outperformed alternatives (e.g., Micro-turbines or Lead-acid batteries).
Why Li-ion over Lead-acid? Despite higher upfront costs, Li-ion's longer lifespan (15 years vs. 5 years), higher round-trip efficiency (90% vs. 75%), and deeper depth of discharge resulted in lower lifecycle costs and emissions.
Why Diesel over Micro-turbine? Although micro-turbines have lower emissions per kWh, their significantly higher capital cost and the fact that they run infrequently (only as backup) made the Diesel Engine more cost-effective in this specific hybrid context.

B. Sizing and Capacity

Renewable Dominance: The optimal system achieved >95% renewable energy penetration.
Oversizing: To ensure reliability without a macrogrid, the system required significant oversizing of RES capacity (~105% of average daily load).
Battery Sizing: The optimal battery capacity was ~52.5% of the total daily energy consumption. Interestingly, the study found that under optimal dispatch, the battery cycled only ~69–85 times per year, implying a replacement period of 60+ years, meaning the battery would likely never need replacement during the 25-year system life.

C. Sensitivity Analysis

Fuel Prices: Counter-intuitively, higher diesel prices improved the overall system performance (lower weighted cost) by incentivizing larger battery storage and higher renewable penetration, reducing reliance on the expensive backup generator.
DG Rating: A backup generator rated at 16 kW (oversized relative to peak load) ensured 100% reliability. However, reducing the DG rating to 8 kW (allowing a 1% load deficiency) significantly lowered costs and emissions, suggesting a trade-off where slight reliability relaxation yields major economic benefits.
Load Profiles: Systems performed significantly better with flatter load profiles. Load shifting (Demand Response) to align consumption with solar/wind availability reduced costs and emissions by up to 50% compared to peakier profiles.

D. Robustness to Uncertainty

The system is highly sensitive to wind speed variability. Low wind periods forced the diesel generator to run continuously, drastically increasing costs and emissions.
The system is more sensitive to supply-side uncertainty (weather) than demand-side fluctuations.

5. Significance and Implications

Policy and Planning: The study provides a quantitative basis for policymakers in SSA to prioritize microgrids over grid extension for remote areas. It demonstrates that hybrid renewable systems are not only environmentally superior but also economically viable when sized correctly.
Design Guidelines: It highlights that "one-size-fits-all" sizing is insufficient. Designers must account for local fuel prices, specific climate patterns (wind vs. solar dominance), and the trade-off between reliability (DG size) and cost.
Future Directions: The paper identifies the need for Demand-Side Management (DSM) and Model Predictive Control (MPC) to further mitigate the risks of renewable intermittency. It also suggests that future models should incorporate thermal loads (CHP) and alternative storage (pumped hydro) to further diversify options for African communities.

In conclusion, this paper establishes that Particle Swarm Optimization is the most effective tool for designing islanded microgrids in Sub-Saharan Africa. It proves that a Solar-Wind-Li-ion-Diesel hybrid system, optimized with multiobjective criteria, offers the most robust, affordable, and sustainable path to electrification for remote communities, provided that operational strategies are flexible enough to handle renewable intermittency.