⚡ electrical engineering

AC-OPF Feasibility Analysis and Sensitivity-Guided Capacitor Placement in a High-PV Islanded Microgrid

This paper proposes a sensitivity-guided capacitor placement strategy within a digital twin framework for a high-PV islanded microgrid, demonstrating that targeted reactive support upgrades can restore full load service and offer a quantifiable economic trade-off against load shedding costs.

Original authors: Aaron Jones, Marija Ilic

Published 2026-04-17

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Aaron Jones, Marija Ilic

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 a small, isolated island town that generates its own electricity using a mix of diesel generators and a massive array of solar panels. This town is "islanded," meaning it's cut off from the main power grid; it has to balance its own books, moment by moment.

This paper is like a stress test and a repair manual for that town's power system when the sun is shining very brightly (high solar penetration). The researchers used a "Digital Twin"—a perfect, virtual copy of the town's power grid—to run simulations and figure out how to keep the lights on without burning out the system.

Here is the story of what they did, broken down into simple concepts:

1. The Problem: The "Too Much Sun" Paradox

Usually, we think having more solar power is always good. But in this isolated town, too much solar can actually cause a crisis.

The Analogy: Imagine a water pipe system. If you turn on too many faucets (solar panels) at once, but the pipes (wires) aren't strong enough to handle the pressure changes, the whole system can burst or collapse.
The Reality: The solar panels generate electricity (real power), but they also need to help manage the "pressure" (voltage) in the wires. If they don't, the voltage gets too high or too low, and the system becomes "infeasible"—meaning the math says the lights must go out, even if there is plenty of energy available.

2. The Four Scenarios (The "What If" Games)

The researchers ran four different simulations over 47 hours to see how the town would react to different rules:

Case 1: The "Cheap & Cheerful" Mode.
The computer tries to run the town as cheaply as possible. It works fine. The lights stay on, and the cost is low. This is the "baseline" (the normal day).
Case 2: The "Stress Test."
They changed the settings on the solar panels to make them behave poorly (changing their "power factor"). Suddenly, the system started to struggle. The voltage got wobbly, and the computer struggled to find a solution. It was like trying to balance a broom on your finger while someone keeps shaking the floor.
Case 3: The "Cut the Power" Solution.
Since the system was struggling, the computer decided to save the day by turning off some lights (load shedding). It cut power to about 16 MW of demand to keep the rest of the town running.
- The Lesson: You can keep the system stable, but the price is that some people lose power.
Case 4: The "Fix the Pipes" Solution.
Instead of turning off lights, the researchers asked: "Where should we add capacitors?"
- What is a capacitor? Think of it as a shock absorber or a pressure tank for electricity. It helps smooth out the voltage swings so the solar panels don't cause a crash.
- The computer used a special "sensitivity score" (like a heat map) to find the exact spots where adding these shock absorbers would help the most. They installed them virtually, and suddenly, the system ran perfectly again with no lights turned off.

3. The Big Discovery: The "Sensitivity Score"

How did they know where to put the capacitors? They didn't guess.

The Analogy: Imagine a doctor diagnosing a patient. Instead of just saying "the patient is sick," the doctor looks at specific vital signs to see exactly which organ is struggling.
The Method: The computer looked at the "vital signs" of the grid (how much the voltage or power was struggling at each specific street corner). It gave every location a score. The locations with the highest scores were the ones that needed the "shock absorbers" (capacitors) the most.
The Result: They found that putting capacitors in just three specific spots fixed the whole problem.

4. The Economic Decision: Fix It or Cut It?

The paper ends with a very practical business question: Is it cheaper to fix the grid or to cut the power?

The Math: They calculated that by installing the capacitors, they saved 16 MW of power that would have otherwise been cut.
The Price Tag: The cost to install the capacitors was equivalent to about $97 for every Megawatt of power saved.
The Verdict: If the town values its reliability (the cost of having a blackout) at more than $97/MW, then buying the capacitors is a smart investment. If they value reliability less than that, they might just accept the blackouts.

Summary: What Does This Mean for You?

This paper is a roadmap for how to handle the future of green energy.

Solar is great, but tricky: When you have a lot of solar, you can't just look at how much energy you have; you have to look at how stable the system is.
Don't guess, measure: You don't need to guess where to put new equipment. You can use computer models to find the exact "weak spots" and fix them surgically.
Investment vs. Outage: Sometimes, spending money on infrastructure (like capacitors) is cheaper than the cost of losing power. This study gives a clear way to calculate that trade-off.

In short, the authors built a digital simulator to prove that with the right "shock absorbers" placed in the right spots, an island town can run entirely on solar and diesel without ever having to turn off the lights, even on the sunniest days.

1. Problem Statement

The paper addresses the challenge of planning and operating islanded microgrids with high penetration of Distributed Energy Resources (DERs), specifically solar Photovoltaic (PV) systems.

The Core Issue: In islanded mode, voltage stability is governed entirely by local generation and reactive power resources. High PV penetration can lead to AC Optimal Power Flow (AC-OPF) infeasibility, where the system cannot sustain voltages within prescribed bounds despite having sufficient active power.
The Gap: Existing methods often rely on linearized (DC) models or single-objective studies that fail to capture the complex nonlinear interactions between voltage, reactive power, and topology in islanded scenarios. There is a lack of frameworks that bridge operational analysis with infrastructure planning decisions (e.g., where to place capacitors) using a unified economic and reliability perspective.

2. Methodology

The authors utilize a Digital Twin (DT) framework (DyMonDS platform) to simulate a real-world, city-scale islanded microgrid (4 MW peak, 12.47 kV) over a 47-hour time-series horizon. The study employs the AC Extended Optimal Power Flow (AC XOPF) framework, which solves the full nonlinear power-flow equations.

The core of the methodology is a four-case comparative study under a shared high-PV loading scenario:

Case 1 (Economic Baseline): Minimizes generation cost with hard voltage limits. Establishes the baseline economic operating point.
Case 2 (Voltage Stress): Simulates adverse conditions by varying PV power factors. This pushes the system toward its voltage-feasibility boundary to expose sensitivity to reactive power assumptions.
Case 3 (Optimal Load Delivery - OLD): Introduces load shedding as a corrective variable to restore feasibility. This quantifies the reliability penalty (cost of lost load) required to maintain operation without infrastructure upgrades.
Case 4 (Capacitor-Enhanced): Uses sensitivity-guided placement to install shunt capacitors at specific buses. The economic objective is rerun to see if infrastructure investment can restore full load service.

Sensitivity-Guided Placement Strategy:

The authors extract optimization sensitivities from the AC-OPF solutions:
- $OS_Q$ : Marginal impact of reactive power injection at a bus.
- $OS_V$ : Marginal impact of relaxing voltage bounds at a bus.
A Composite Placement Score ( $S_k$ ) is calculated for each bus:
$S_k = w_Q |OS_Q^{(k)}| + w_V |OS_V^{(k)}|$
This score ranks candidate buses for capacitor upgrades.
A Post-Processing Planning Optimization compares the cost of capacitor installation ( $C_{cap}$ ) against the expected Value of Lost Load ($VoLL$) derived from Case 3, enabling a direct economic trade-off analysis.

3. Key Contributions

Comparative AC-OPF Framework: A novel four-case study on a real-world islanded microgrid that demonstrates how different objective functions (economic, stress, load delivery, infrastructure) reveal complementary planning signals.
Composite Sensitivity Score: Development of a weighted score ( $S_k$ ) combining reactive and voltage sensitivities to rank capacitor placement candidates, allowing planners to adjust weights based on specific grid goals.
Economic Planning Bridge: A post-processing optimization that explicitly balances infrastructure investment costs against avoided reliability penalties (VoLL), providing a quantifiable metric for investment justification.
Feasibility Restoration: Demonstration that sensitivity-guided capacitor placement can fully restore time-series feasibility (serving 100% of the load) across the entire horizon, matching the performance of the economic baseline without the need for load shedding.

4. Key Results

The study was conducted on a 4 MW microgrid with two utility interconnection points (disconnected for islanded operation) and three distributed solar units.

Case 1 (Baseline): Successfully served 87.81 MW at a cost of $8,605 with tight voltage regulation (1.016–1.019 p.u.).
Case 2 (Stress): Under altered PV power factors, the system degraded significantly. Voltage mismatch increased 40x, and voltages drifted to 1.043–1.05 p.u., indicating the system was near the feasibility boundary.
Case 3 (Load Shedding): To restore feasibility without hardware changes, the system had to curtail 16.07 MW of load. The cost dropped to $7,045, but reliability was compromised.
Case 4 (Capacitor Enhancement): By installing capacitors at the top-ranked buses (508, 364, 675) identified by the sensitivity score:
- Full load (87.81 MW) was restored.
- Total cost was $8,605.34 (virtually identical to the baseline).
- Voltage profiles were tightened (1.015–1.019 p.u.).
Economic Trade-off: The cost to recover the 16.07 MW of shed load via capacitors was an additional $1,559 over the 47-hour horizon. This equates to $97 per MW of recovered demand.
- Decision Rule: If the local Value of Lost Load (VoLL) exceeds $97/MW, capacitor installation is economically justified.
Sensitivity Consistency: Buses 508 and 364 were consistently ranked as top candidates across all cases. However, the third-ranked bus shifted (675 vs. 783) depending on whether load shedding occurred, proving that a single-objective study would miss critical planning nuances caused by load redistribution.

5. Significance

This paper provides a robust, data-driven workflow for microgrid planning under high DER penetration.

Beyond Hosting Capacity: It moves beyond simple "how much PV can we fit" questions to "how do we manage voltage feasibility dynamically."
Operational-to-Planning Bridge: It effectively translates real-time operational sensitivities (from AC-OPF) into long-term infrastructure investment decisions.
Quantifiable ROI: By framing the decision as a comparison between corrective load shedding and preventive infrastructure, it gives planners a clear, economic threshold ($97/MW in this case) for justifying capacitor upgrades.
Generalizability: The methodology is network-agnostic; while the numerical results are specific to the studied topology, the sensitivity-guided ranking and planning optimization framework can be applied to any islanded microgrid facing similar voltage stability challenges.

In conclusion, the study proves that sensitivity-guided capacitor placement is a highly effective strategy to maintain reliability in high-PV islanded microgrids, offering a cost-effective alternative to load shedding while ensuring full service delivery.