Island Hybrid Mini-Grid: Diesel Efficiency, Storage Sizing, and Cyclone Resilience
An island mini-grid serves 18 MW peak demand and 6 MW minimum demand. The installed fleet is four 2.5 MW diesel engines, 8 MWp fixed-tilt PV, three 1.5 MW wind turbines, and a 2 MW/4 MWh lithium-ion battery. Annual demand is 72 GWh. In the first year, instantaneous renewable penetration reached 45%, but diesel fuel use fell only 18% against the pre-project baseline. Operators report midday PV curtailment, reverse power flow at the diesel plant, engines running at 25-35% load for long periods, and a specific fuel consumption of 0.30 L/kWh against the 0.21 L/kWh guaranteed at 75% load. The battery cycles 1.8 times per day and is often at its state-of-charge limit by late afternoon. The island is exposed to category 4 cyclones; the PV arrays have stow capability, but the wind turbines are rated to IEC class III and the battery room is at 2.5 m above sea level. The utility must decide whether to add 4 MWh of battery, 2 MWp of PV, or a fourth wind turbine, and how to improve resilience without worsening diesel efficiency.
1. Reconcile the 45% instantaneous renewable penetration with the 18% fuel savings. Calculate the fuel and cost penalty of low-load diesel operation, assuming 12 GWh of diesel generation and diesel at $0.85/L.
Read the worked answer
45% is an instantaneous share at certain hours, not an annual energy share. Annual renewable generation may be only about 28-32% because wind and PV are curtailed and the battery cannot move enough energy. Diesel savings are further eroded by minimum-load operation. With 12 GWh of diesel generation, fuel at 0.30 L/kWh is 3.6 million L; at the guaranteed 0.21 L/kWh it would be 2.52 million L. The penalty is 1.08 million L, or about $918,000/year at $0.85/L. That is the cost of running engines at 25-35% load to absorb variability. The battery is cycling 1.8 times/day and is often full by late afternoon, so midday PV is curtailed while evening peak still runs diesels. The fix is not simply more renewables; it is dispatch control that reserves battery headroom for midday charging, sequences engines so one runs near 70-80% load, and treats curtailment as a measured KPI.
2. Evaluate the three expansion options against the constraint that at least one diesel engine must remain above 40% load. Which would you choose and why?
Read the worked answer
Add 4 MWh of battery first, but only with a dispatch and sizing review. The battery addresses the binding constraints: it absorbs midday PV that is currently curtailed and can keep a diesel engine off or above 40% load during transitions. However, 4 MWh is only about 20-30 minutes at 8-10 MW of net load, so it will not solve the evening peak alone. A 2 MWp PV addition increases midday surplus and would likely increase curtailment unless storage is expanded first; it also adds little firm capacity. A fourth wind turbine adds night generation and complements PV, but the existing turbines are IEC class III, and a category 4 cyclone site may require class I/II or a validated site-specific load case; that is a resilience and warranty issue, not just a cost issue. Recommendation: add 4-6 MWh of battery with grid-forming capability and revised dispatch, defer PV, and procure wind only if the turbine can be certified for the site and the network can accept the night-time reverse flow. Reassess after one year of measured curtailment and engine loading.
3. Recommend a resilience package for category 4 cyclones, considering the battery room elevation and wind turbine class. How would you justify the cost?
Read the worked answer
Resilience package: relocate or elevate the battery room above the 2.5 m level, ideally above the 100-year storm surge plus freeboard; add flood barriers, sump pumps, fire detection appropriate for lithium-ion, and waterproof cable entries. For PV, keep stow capability and add rapid shutdown and spare inverters; for wind, either replace with cyclone-rated class I/II turbines or specify a site-specific extreme wind load case and a curtailment/stow strategy. Add at least 72 hours of diesel fuel, a black-start capable battery or generator, and priority circuits for water, clinic and communications. Justify with expected annual loss: estimate outage frequency and duration, value lost load and spoiled product, and compare with the incremental cost. If the incremental resilience cost is $1.5-2.5 million and avoids a 5-day outage costing $3-5 million with 20% annual probability, the expected benefit is $0.6-1.0 million per year, which is easy to justify; also check insurance and regulatory cost recovery.