Summary: Chilean solar projects in the Atacama face a dual trap: ultra-low tariff auctions force Atacama site selection despite transmission bottlenecks. When midday generation peaks and transmission lines saturate, the grid operator (CEN) issues mandatory curtailment. Projects generate 0 MW but still owe power to off-takers under standard delivery-node PPAs. The CEN cannot legally order price spikes, so fossil plants ramp in central Santiago (Quillota node), pushing spot prices to $80/MWh. Developers must buy spot at $80 to sell contracted power at $45, creating a negative margin per megawatt. A 150MW project modeled at 9.5% IRR with 10% curtailment haircut realizes -2.5% when curtailment clusters during peak delivery periods.
Chilean Grid Physics Create Synthetic Short Positions for Solar Developers
Chilean developers treat long-term PPAs as comprehensive hedges against spot market volatility. They assume grid connections at world-class solar nodes automatically translate to grid evacuation capacity. They model PPA strike prices as hard revenue floor and cost-plus returns.
This assumption ignores Chile's linear grid topology: The "Nodal Short" Trap. In a geographically linear network, signing a delivery-node PPA without holding physical transmission bypass capacity functionally creates a synthetic short position against the grid operator. When curtailment occurs, your asset generates nothing, but your financial obligation remains.
The breakdown occurs mechanically through the National Electrical Coordinator's (CEN) automated dispatch logic. When north-south transmission corridors saturate during peak Atacama solar hours (9 AM-5 PM), the CEN issues mandatory curtailment orders. Generation drops to zero. However, delivery-node PPAs still require the developer to supply power to Santiago-based off-takers. Because northern supply is physically trapped, Santiago must rely on localized gas peakers, spiking Quillota spot prices to $80/MWh. The curtailed developer must buy spot at $80/MWh to sell contracted power at $45/MWh, creating a -$35/MWh margin.
Consider a 150MW solar plant at the Crucero node with a $45/MWh delivery-node PPA to a Santiago off-taker. At 14:00, transmission saturates. CEN orders 100% curtailment. The plant generates 0 MW. The developer still owes 150 MWh for that hour. Santiago's fossil peakers spike prices to $80/MWh. The developer must purchase 150 MWh at $80 to honor the contract at $45, losing $5.25M that hour.
An 8760-hour model applying a flat 10% volumetric curtailment haircut projects 9.5% IRR. In reality, because curtailment clusters during peak delivery obligations, the unhedged replacement costs drain reserve accounts in months. Realized IRR collapses to -2.5%, breaking DSCR covenants and forcing distress.
Current prefeasibility tools model uniform grids. They apply static zonal pricing and volumetric haircuts, failing to simulate sub-hourly thermal bottlenecks and the catastrophic financial penalty of owing power at a premium node while isolated at a zero-price curtailed node.
Investors must immediately reject standalone solar in the Atacama. Bankability now strictly requires Battery Energy Storage Systems (BESS) sized to shift un-curtailed evening power into the PPA delivery window, functioning as a mandatory bypass valve rather than a revenue optimization tool.
Bottom line: In a geographically constrained grid, exceptional sunshine is not an asset; the only bankable asset is unimpeded transmission access.
Calculating sub-hourly nodal price decoupling, modeling basis risk between Chilean injection and withdrawal nodes, and sizing BESS to offset replacement power costs requires dynamic grid simulation. Static uniform-grid spreadsheets cannot capture the lethal interaction between physical curtailment and financial settlement obligations. Accurately sizing defensive storage requires preFeasibility environments that model nodal price decoupling against real-time grid constraints.