Ex-Ante Projection — BESS Under Construction— 6,357 MW

Heatmap = net EF for any hypothetical BESS of (Power, Capacity) size run against the 2024 SEN baseline. Blue line = the 74-project pipeline treated as a single cumulative BESS: each point is the whole portfolio up to that date, not individual project credits. The trajectory advances by declared COD date.

Curtailment saturation. As the aggregated BESS grows it absorbs more of the renewable curtailment pool. Once that pool is exhausted, additional capacity must charge from thermal generation — which is why the initial EF descent flattens out around 2027 instead of collapsing to zero.

Frozen operating curves. Charge and discharge profiles are drawn from a statistical model fitted to the 41 real SEN BESS currently operating (2025 hourly data). They remain unchanged across all projected sizes — no future re-optimisation (e.g. shifting to wind curtailment hours) is assumed, which keeps the projection conservative.

All BESS are in the merit order. Existing, pipeline (74 under construction) and the project under evaluation all participate in the dispatch order. None receives preferential treatment; all have zero variable cost and compete hour by hour in the marginal stack based on variable costs declared to the Chilean National Electric Coordinator (CEN).

Simultaneous injection does not duplicate credit. When multiple BESS discharge concurrently, a tCO₂ avoided is credited only to the BESS that effectively displaced it in the simulated dispatch.

No systematic gas-vs-coal bias. The model identifies the exact marginal unit hourly using the variable costs reported to the Coordinator. Point errors (e.g. confusing gas with coal) are rare and symmetric in both directions, producing no aggregate over- or under-estimation.

EF does not collapse to zero, even with the full pipeline. In the extreme scenario of all 74 BESS discharging simultaneously (~6.4 GW), the aggregate net EF stabilises at ~0.24 tCO₂/MWh by 2027, with annual RED of ~1.0 MtCO₂/yr. Significant headroom remains before thermal displacement saturates.

Every ambiguity resolved towards fewer credits. The model uses the current profiles of the 41 real SEN BESS (fitted to 2025 hourly data) without optimising for future wind curtailment. Applies DORd with DIF DESC + EF DESC sort (expensive/dirty units displaced first), and filters DPP=0 and EGEMd=0 units.

Ex-post resolves ex-ante uncertainty. This chart is a conservative lower bound. When ITMO calculations run on actual ex-post Coordinator dispatch data, the exact hourly marginal unit per individual BESS is known — without speculation about prioritisation or technology confusion.