Skip to main content

Commercial RPM decision log

Decisions about the Commercial RPM model, newest first. When the team decides something, add a row at the top. Don't rewrite old rows: if a decision changes, add a new row and say which one it replaces. Current terms in one place: NFE Commercial Reliable Power Microgrids. Owner: Aaron Tushabe.

Date Decision Why Who / where
21 Oct 2026 Fundraising target (provisional): US$500,000 of total programme capital to target ten new RPMs over about 18 months: seven Commercial and three Residential (70% commercial, within the 60–80% range). US$550,000 if we want room for a fourth residential site. Replaces the US$650,000 working estimate from earlier the same evening. Sized to deploy about ten microgrids with a commercial majority. The first estimate used Worship Harvest's scale for every commercial site. Aaron set the scope; budget worked up in his ChatGPT session. See Funding target
21 Oct 2026 Planning assumption: Worship Harvest (~120 kWh) is one of the larger Commercial RPMs, not the typical size. Budget most commercial sites as smaller (30–40 kWh, ~US$15,000 installed) or medium (60–80 kWh, ~US$25,000). Aaron's expectation of the commercial pipeline. Aaron
21 Oct 2026 The funders' pitch will present Commercial RPMs as NFE's near-term expansion priority alongside Residential RPMs, with Worship Harvest kept anonymous. Draft v0.1 is in review. The web page stays unchanged until the fundraising amount and deployment milestones are set. NFE's growth plan has shifted toward commercial customers. Aaron
6 Oct 2026 Price: ERA Code 10.2 time-of-use tariffs, with applicable taxes shown separately. Reliability is included; no separate reliability fee. "Reliability baked in" is the pitch. A fee would change the offer. Aaron, after a teammate's review of the Worship Harvest proposal
6 Oct 2026 Minimum monthly payment: from commissioning, the customer pays the greater of the actual metered energy charge or 50% of the reference monthly energy charge agreed after Phase 1 submetering. The baseline kWh is fixed in the contract and tariff changes adjust its value. It replaces a lower bill and is never added on top. Service-performance credits apply if NFE fails to deliver. Replaces "no hard minimum" (18 Sep). Pure pay-as-you-use left NFE with all the volume risk for 15 years. A 50% floor gives lenders some revenue protection, while the customer's use can fall a long way before it applies. Aaron
6 Oct 2026 Utility account: NFE prefers to manage and pay the customer's UEDCL bill and issue one consolidated energy bill. The customer decides during Phase 1. Formal account transfer only if needed. If the customer keeps paying UEDCL directly, agree a reconciliation so they aren't charged twice. One bill is simpler for the customer, and NFE has to be the billing layer to capture the solar margin. Aaron
6 Oct 2026 Approvals: the ERA requirement is a certificate of registration for the microgrid. Replaces the broader "confirm licensing, registration or exemption" wording. Sets out the one approval NFE needs. Aaron
6 Oct 2026 Ownership wording: customer documents say "Optional purchase from year 5 (Ownership stage), subject to agreed terms", not "Phase 4". "Phase 4" clashed with step 4 ("Install and commission") of the delivery plan. Aaron, on a teammate's suggestion
6 Oct 2026 Savings illustrations use the customer's historical weighted Code 10.2 tariff (Worship Harvest: UGX 567.1/kWh) plus assumed 18% VAT (UGX 669.2/kWh). The UGX 800/kWh figure is dropped. Sunday demand will come from Phase 1 submetering. The proposal priced energy at the tariff but illustrated savings at UGX 800, which nothing explained. Aaron, after a teammate's review
2 Oct 2026 NFE's roadmap will include more commercial than residential microgrids. — Aaron
2 Oct 2026 Phase 1 (load and outage monitoring) normally lasts 3 months. Installation follows 3–6 months after the load and outage profiles are confirmed. Size each microgrid from measured data. Aaron
2 Oct 2026 NFE's platform is called Microgrid Manager in all documents, not MicroPowerManager. NFE now uses Microgrid Manager. Aaron
1 Oct 2026 The website Products section shows two products: Residential RPM (50–500 residents) and Commercial RPM (at least 2 MWh/month). Customer-facing pages leave out internal economics and financing. Headline: "Energy as a Service, not another equipment project". Prospects need to know whether they fit and what to prepare before contacting us. Aaron
19 Sep 2026 Pursue Worship Harvest, but treat the financial and engineering figures as unvalidated until reviewed. Peak demand is the biggest gap: meter after the grid/generator changeover, capture generator and event loads, and add avoided diesel to the model. Worship Harvest runs a 200 kVA generator on Sundays and for events, so its bills understate both demand and the value at stake. Team, check-in
18–19 Sep 2026 Screen commercial leads at ≥2 MWh/month (≥3 MWh is stronger). Preliminary. Modelled break-even was about 1.5 MWh/month; 2 MWh leaves room for downside cases. Aaron's model; team check-in
18 Sep 2026 Commercial Phase 2 = a battery for reliability + enough solar to finance it. At tariff parity, solar is what pays for the reliability. Aaron's model; confirmed at the 19 Sep check-in
18 Sep 2026 Fund the battery and inverter replacement reserve explicitly in the model (lenders won't accept a generic % of CAPEX). Present Worship Harvest CAPEX as a range (UGX 120–145m; UGX 140m in the model), not as "UGX 143m". UGX 143m was a conservative modelling number, not a quote. Aaron
18 Sep 2026 Commercial RPMs keep NFE's Energy as a Service model: per-kWh billing at utility-tariff parity, NFE owns and runs the equipment, four phases with an ownership option. Not an equipment lease like Sawa's, but borrow Sawa's "set it and forget it" risk allocation and predefined buyout. Keeps the offer simple: pay for power, not power equipment. Aaron
12 Sep 2026 Explore a commercial/industrial model as an adjacent opportunity, without distracting from the residential/community strategy. Superseded by the 2 Oct shift toward commercial. A business with a predictable UGX 4–5m monthly bill could fund a system from its existing energy spend. Team, check-in

Source: Aaron's ChatGPT working session "Commercial RPM Modeling" (18 Sep–21 Oct 2026) and the linked check-ins.

No comments to display

Back to top