Nanogrid Market Lab
Build a village from appliances up, choose how neighbours trade their surplus, and see the price, the tariff and the CAPEX before any hardware ships.
Market logic follows Bertram, Fuchs, Banuls Ramirez, Crespo del Granado & Balderrama, “Local electricity market designs for interconnected nanogrids,” Journal of Cleaner Production 449 (2024) 141786. Demand is built bottom-up from appliances at 15-minute resolution — see methodology.
THE VILLAGE MARKET
one representative dayEach node is one nanogrid: a single solar panel and battery shared by four to six households, run by a local operator who trades on their behalf. Trading happens between nanogrids — never inside one.
AVERAGE TRADING PRICE THROUGH THE DAY
€/kWhHOUSEHOLD LOAD BY APPLIANCE
ENERGY FLOWS & BATTERY
MARKET DESIGN COMPARISON
BOTTOM-UP END-USER TARIFF
€/kWh by power subscriptionAs a percentage of the entry-level tariff.
NANOGRID FLEET
Sites are seeded from the paper's 18 nanogrids in the Diana Region, Madagascar. Appliance level scales the household mix for that site — it is what makes some sites net exporters and others net importers. Daily demand is derived, not typed.
- Demand is built from appliances. Every 15 minutes, each appliance draws its rated power times a duty profile — lamps and TVs in the evening, fridges around the clock with a hot-afternoon bump, mills and pumps in the working day. A site's load is its household count × appliance level × the household mix, plus its own productive equipment. Nothing about the load shape is assumed; it emerges from what people own.
- Service levels. "Lighting only" reproduces the paper's measured fleet at about 40 Wh per household per day. Tier 2 and Tier 3 follow the Multi-Tier Framework's ladder. Raising the tier without resizing the hardware is what produces unserved energy — which is the point.
- Prices. A seller asks its own LCOE: panels alone in daylight, panels plus battery replacements after dark. A buyer bids the retail tariff its households already pay, as a ceiling. Central clearing settles at the asking price of the last site needed and everyone pays that; bilateral splits the difference on each matched pair, so the buyer's tariff pulls prices up. That is why bilateral runs dearer without being better at finding energy.
- Unserved energy is demand that the site's own panel, its own battery and the market together could not meet. It is the reliability number a financier asks for first.
- Line losses are a flat percentage per 100 m — enough to make distance cost something without turning this into a cable-sizing exercise. Operator margin is the franchisee's revenue term from Eq. 5, added on top of cost recovery.
- Non-residential loads are ESMAP field data. Shops, schools, water pumps, health clinics and flour mills use the measured hourly profiles from Figure 2.26 of Mini Grids for Half a Billion People (ESMAP / World Bank, source VIA 2021) rather than invented duty curves — a shop draws 78 W through the working day and 1.0 kWh/day; a flour mill peaks at 4 kW and eats 47.8 kWh/day, more than this whole cluster of 88 homes consumes. That contrast is worth seeing: a mill is not an appliance you bolt onto a nanogrid, it is a reason to build a bigger system.
- Benchmarks. Cost per connection, CAPEX per kW, load factor and fleet LCOE are shown against ESMAP's published figures ($1,000–2,100 per connection; $3,908/kW falling below $3,000 by 2030; 22% load factor baseline rising to 40% with productive use, cutting LCOE from $0.55 to $0.42/kWh). ESMAP's average residential profile (303 Wh/household/day, 30 W peak) is drawn on the household load chart so an appliance build can be checked against it.
- Still assumed: household appliance duty patterns are typical, not metered from a specific village; weather is a random daily factor; agents bid honestly at cost with no strategic withholding.