The market Folio Grid models.

01. What the engine forecasts

Folio Grid sits where the underwriter sits. Its job is to take a proposed or existing utility-scale battery or hybrid renewable project and produce a bankable merchant revenue stack, a grid-connection risk read, and a counterparty view that a real investment committee can sign on. The engine is an independent model. It starts from first principles on what the electricity market ought to deliver, not from a vendor's pro forma.

The stack is built on two named models. Lattice simulates the wholesale dispatch and the transmission network: prices, flows, losses, congestion. Reservoir sits on top of Lattice's outputs and optimises the dispatch of the specific battery or hybrid asset: charge-discharge scheduling across wholesale, ancillary services, and contracted revenue layers under realistic foresight and engineering limits. Lattice tells you what the market is doing at the connection point. Reservoir tells you how many dollars your asset can actually extract from that market across the asset's life.

Primary markets in scope are the Australian National Electricity Market, the US ERCOT market, and Gulf Cooperation Council markets, in that order of depth. NEM is regional-zonal, ERCOT is nodal with locational marginal pricing, and GCC markets sit between capacity-based and bilateral. The two models carry the design differences explicitly rather than collapsing them into one abstraction.

Forecast horizons run from the next five-minute trading interval to fifty years out. Time granularity matches the decision at hand: half-hourly for merchant revenue, five-minute for ancillary service stacking, annual for scenario planning.

02. The five market surfaces Folio Grid reads

Every IC-grade diligence pack rests on five distinct market surfaces. Four of them feed Lattice. The fifth, battery dispatch, is Reservoir's domain. The five outputs combine into a single bankable merchant stack the investor can stress-test.

Wholesale electricity market (Lattice). The engine simulates the dispatch of the wholesale market, forecasts prices at half-hourly resolution, and produces generation and reliability outputs across the horizon. Inputs capture market rules, regulatory changes, and portfolio bidding behaviour. Outputs feed both the revenue stack and the reliability risk view.

Network power flows and loss factors (Lattice). The engine models the transmission network to quantify marginal loss factors and distribution loss factors for the proposed point of connection. Loss factors cut directly into generator revenue. The same model identifies congestion risk at the connection point and surfaces it as a curtailment expectation rather than an unmodelled tail.

Large-scale Generation Certificate market (Lattice). The engine analyses LGC supply and demand, the position of retailers under the Renewable Energy Target, and the required new-build revenue to sustain the scheme. LGC prices feed the revenue stack for projects that qualify.

Wind and solar yield (Lattice). The engine simulates generation from the proposed wind or solar resource at half-hourly resolution using weather histories from the Bureau of Meteorology in Australia and equivalent public records in each market. Yield models are calibrated to site-specific wind speeds and solar irradiance. A three percent P50 shift on a hybrid asset's renewable tail can move the merchant stack by more than the battery's contracted spread.

Battery and hybrid asset dispatch (Reservoir). The engine runs the asset-level optimisation: state-of-charge, cycling limits, degradation assumptions, round-trip efficiency, contract overlays, ancillary-service participation, and imperfect foresight of future prices. Outputs feed the revenue stack and the bankable IC memo.

03. Lattice: the wholesale market and network model

Lattice emulates the dispatch engine that the market operator uses to clear the market. In gross-pool markets like the NEM, every dispatchable generator bids its output into each five-minute trading interval at a price. The bids form a stack ordered from cheapest to most expensive. Price is set by the marginal generator that clears the forecast demand for that interval. In nodal markets like ERCOT, the same logic runs node by node and the locational marginal price reflects both the marginal cost and the network constraints at that node.

Lattice captures both regimes. It runs short-run marginal cost bidding and strategic portfolio bidding, generator outages and maintenance cycles, renewable intermittency, day-ahead versus real-time market splits, transmission congestion, ramp-rate constraints, and the price spikes that follow when any of these binds. It runs iteratively until the dispatch reaches an equilibrium that matches portfolio profit objectives. Bids can go negative, and Lattice captures that explicitly: static cost-stack shortcuts miss negative-price hours and systematically understate curtailment risk for renewables.

The network layer inside Lattice models power flows on the actual transmission topology. It forecasts marginal loss factors and distribution loss factors for the project's connection point out to the horizon, not just the current-year value. It quantifies congestion-driven curtailment at the connection point and surfaces augmentation options with indicative cost and lead time. In nodal markets, Lattice produces the LMP that flows straight into Reservoir; in zonal markets it produces the regional reference node price plus the connection-point yield adjustment. The combination is what the engine calls the connection-point yield.

04. Reservoir: the BESS and hybrid dispatch model

Reservoir takes Lattice's price series and turns them into an asset-level dispatch schedule that respects how a battery actually operates. It runs for standalone BESS, co-located BESS plus renewable assets, and renewable projects with a battery-backed hybrid PPA. The model is calibrated against historical operating data so the dispatch it recommends reflects what real operators achieve, not what a theoretical optimiser would extract under perfect foresight.

The optimisation respects the engineering and commercial constraints that move dispatch in practice. State-of-charge bounds and round-trip efficiency. Cycling limits and warranty envelopes. Degradation curves across the asset's life. Minimum dispatch quantities on ancillary markets. Regulatory constraints on import or export at the connection point. Contracted revenue overlays from tolling agreements, floor-price contracts, or hybrid PPAs that cap one revenue stream in exchange for another.

Reservoir trades imperfect foresight explicitly. A real battery operator does not know tomorrow's prices with certainty. Reservoir's scheduler operates on forward price signals with a realistic uncertainty band, not a retrospective optimiser that cherry-picks the top intervals. The output is a half-hourly dispatch profile across the asset's life, decomposed by revenue stream: energy arbitrage, frequency response, contingency FCAS, inertia and system services where applicable, capacity payments where the market provides them, and hybrid PPA net revenues. That decomposition is what makes the revenue stack bankable instead of a single aggregate line.

05. What drives the spot price

A spot price in any given interval is a function of six inputs. Lattice models each of them explicitly and exposes them as scenario levers in the final memo. In nodal markets the same drivers apply per-node with the addition of network topology and congestion.

Demand. Average load versus peak load, and the shape of the load duration curve across the year. Mis-shaped demand forecasts are the single largest source of error in merchant revenue modelling.

Commodity prices. Coal, carbon, and gas feed directly into the cost of the marginal generator, especially in the transition period while fossil plant still sets the price in many intervals.

Capacity mix. Total installed capacity by technology, the fixed-cost recovery requirements of each plant, availability factors, and the rate at which new plant is entering or leaving the market.

Market behaviour. Portfolio bidding strategy, scarcity pricing behaviour, the policy and environment around compliance markets, and the risk profile of the trading entities in the market.

Cross-border exchange and congestion. Interconnector flows between NEM regions; limited cross-regional transfer capacity in ERCOT; nodal congestion that moves LMPs apart across a small geographic area. Flexibility of the interconnector during congested intervals changes the price footprint materially.

CO2 pricing and the carbon policy trajectory. Efficiency and profitability of each thermal unit depend on the carbon cost assumption used. The engine runs a scenario on this rather than baking a single path.

06. Sample engagement shapes

Folio Grid is built to serve a handful of distinct engagement shapes. The patterns below are indicative rather than exhaustive: the engine runs the same two models underneath each of them, but the deliverable, cadence, and pricing envelope differ by shape.

Single-asset bankable IC memo. A developer, sponsor, or PE fund is underwriting one utility-scale BESS, co-located BESS plus PV or wind, or hybrid renewable project and needs a memo an investment committee can sign against. Target turnaround: one week for a NEM asset at the methodology's current depth. Output: the bankable merchant revenue stack decomposed by revenue stream, the connection-point yield read, the LGC strip where eligible, a counterparty view on the offtake or tolling counterparty, a debt-sizing sensitivity envelope, and a signed-off IC-grade narrative.

Portfolio revaluation. An owner of a BESS or hybrid portfolio (pre-IPO, pre-refinancing, or quarterly-valuation cadence) needs a mark-to-model of the fleet. Lattice runs a shared market outlook, Reservoir runs asset-by-asset dispatch with the current contract overlays, and the output is a portfolio-level revenue forecast distribution with asset-level contribution analysis. Delivered as a recurring subscription with quarterly refreshes.

Offtake and hedge structuring. A sponsor negotiating a contract, floor, or hybrid PPA needs to know the value of the contracted layer versus the merchant tail under realistic dispatch. Reservoir produces the contracted-versus-merchant split under a candidate contract structure; repeated runs across alternative structures produce the negotiation envelope. Used by both sides of the table.

Greenfield siting and interconnection screen. A developer evaluating a set of candidate connection points against the current network plan needs a quick read on MLF trajectory, congestion risk, and relative merchant upside. Lattice's network layer runs the candidates in parallel and produces a comparison-grade shortlist with indicative augmentation cost and lead time for each. Typically a one-day turnaround per candidate shortlist.

Lender-grade independent market opinion. A lender or lender's adviser needs an independent merchant-market view on a deal they are pricing. Folio Grid produces a lender-grade market opinion that sits alongside the independent engineer's technical report. The deliverable is shorter than a full IC memo and is scoped to the risk questions the credit committee is actually pricing.

07. The LGC strip

For eligible projects, Large-scale Generation Certificates are a separate revenue layer. Each MWh of accredited renewable generation produces one LGC, which retailers buy to meet their Renewable Energy Target obligation. LGC pricing is driven by the retailer supply-demand balance, the build pipeline of accredited projects, the banking behaviour of large market participants, and the policy view on what happens to the target after its current end date.

Folio Grid forecasts LGC prices on a required-revenue basis. The model calculates what the LGC price needs to be to bring forward the marginal new project, then reconciles that against announced build pipeline and current banked volumes. The output is not one LGC price. It is a forecast distribution with a central path and a low-supply stress case, because LGC price is structurally thin and sensitive to a handful of large portfolio decisions.

08. Forecast horizons

Different decisions need different horizons. Folio Grid is built to run at the granularity each decision actually requires.

One day ahead and the next five minutes. For ancillary service modelling, frequency response revenue, and operational dispatch. The engine uses the five-minute or thirty-minute resolution that the market itself clears at.

One year to ten years. For merchant revenue underwriting, hedge design, and offtake contract negotiation. The half-hourly dispatch model runs forward with an evolving cost stack, capacity entry and exit, and portfolio bidding that adapts to the scenario assumptions.

Ten years to fifty years. For scenario planning, policy impact assessment, and long-dated hybrid asset valuation. The engine drops to annual granularity and operates on capacity mix assumptions rather than bid-level detail, but retains the full set of scenario levers so IC committees can see the envelope, not a false central estimate.

09. The market context Folio Grid carries

Any credible model has to know the ruleset it is modelling. Folio Grid's runtime carries an operating picture of the NEM, ERCOT, and GCC markets. At minimum this includes the rule-maker and its amendment cadence, the market operator and its dispatch engine, the market regulator and the appeals path, and the full set of market participants with their distinct roles: generators, retailers, network operators, traders, investors, industrial loads, and general market customers.

The picture matters because rule changes land frequently and most of them shift the revenue envelope in ways a static pro forma cannot capture. The engine has a named owner for tracking each market's rules and a change-log that feeds into the scenario layer.

10. What changes when the market decentralises

The direction of travel is well understood. Power and utilities are moving from a centralised generation model to a decentralised one built on three overlapping forces: decarbonisation, digitalisation, and decentralisation. Distributed generation, distributed storage, virtual power plants, microgrids, electric vehicles as mobile storage, and smart metering at scale all push revenue streams off the bulk system and toward the connection point.

Folio Grid's methodology accounts for this trajectory explicitly. Wholesale merchant revenue is one envelope, and the engine also prices a separate envelope for co-location revenue (where the asset serves a host customer or a virtual power plant aggregator) and for behind-the-meter services. Projects that can access both sides of that boundary carry a different risk profile than projects anchored only to the wholesale pool. The model surfaces that distinction rather than collapsing it.

None of this changes the core diligence question an investor is asking. It changes what the answer has to be made of.