Research Overview

Best Answer

Over 24 months of SEM day-ahead data (Jan 2024 – Dec 2025, 17,567 hourly observations across 732 complete days), a 4-hour battery achieves a median daily spread of EUR 61.7/MWh after 85% round-trip efficiency losses, with a mean of EUR 75.7/MWh. The middle 50% of days fall between EUR 38.7 and EUR 97.1/MWh (P90 = EUR 143.3, P95 = EUR 170.2). 100% of days in the historical period were profitable before fixed costs.

Revenue is moderately concentrated: the top 25% of days generate 49% of revenue, but even the bottom half of days are profitable (median day earns ~EUR 12k for a 50 MW / 200 MWh system). There is strong seasonality — winter months (Oct–Jan) show median spreads of EUR 76–141/MWh, while summer (Jun–Aug) ranges EUR 31–79/MWh. Weekdays average 24% higher spreads than weekends (EUR 81.2 vs EUR 61.9/MWh). 16 days (2.2%) had negative prices.

Key Assumptions

• Historical spreads persist into the future — forward scenario modelling shows significant compression: Celtic Interconnector alone cuts median spreads 35%, +2 GW BESS fleet compresses 38%, and the combined 2030 scenario (all factors) shows an 81% median decline to EUR 11.9/MWh
• Perfect daily cycling is achievable — in practice, forecast errors and market timing reduce capture rate
• Forward scenarios are based on our own peak-capping and fleet-competition models, not published third-party estimates

What Could Flip This — Modelled Scenario Impacts (4hr battery)

• Celtic Interconnector alone (2028): median spread drops from EUR 61.7 to EUR 42.2/MWh — 35% compression from peak-capping effect
• +2 GW BESS fleet competition: median drops to EUR 37.1/MWh — 38% compression
• Combined 2028 (Celtic + wind + BESS): median EUR 28.7/MWh, mean EUR 34.8/MWh, 96.2% of days profitable — 54% decline from current
• Combined 2030 (all factors): median EUR 11.9/MWh, mean EUR 17.5/MWh, 87.0% of days profitable — 81% decline, revenue falls to ~EUR 27k/MW/yr vs EUR 111k currently
• +30% data centre demand growth: partially offsets compression, adding ~5% back to spreads (median EUR 58.5 vs EUR 61.7 baseline)
• +5 GW wind alone (2030): barely changes spreads — only 3% decline (median EUR 60.6/MWh)

Sources: Energy-Charts API / ENTSO-E (24-month hourly data, Jan 2024 – Dec 2025), forward scenarios modelled from empirical base

Best Answer

For transmission-connected BESS, the round-trip cost (excluding the energy purchase price) is approximately EUR 48–66/MWh. The biggest component is the D-TUoS System Services Charge at EUR 26–29/MWh on every MWh imported, even if it's immediately re-exported. This is widely regarded as unfair "double-charging".

Fee Breakdown (Published Rates)

• D-TUoS Network Transfer: EUR 4.09/MWh Solid
• D-TUoS System Services: EUR 26.47–29.12/MWh Solid
• D-TUoS Network Capacity: EUR 2,010/MW/month Solid
• Market Operator: ~EUR 0.61/MWh Solid
• Imperfections: EUR 19.93/MWh (supplier-levied; how this applies to BESS self-supply is unclear) Handwavy
• Efficiency loss: ~15–20% of purchase price Reasonable

What Could Flip This

• System Services charge reform — ESI/ECA propose exempting BESS. If adopted, round-trip cost drops to ~EUR 18–22/MWh, massively improving viability. CRU has NOT adopted this yet.

Sources: EirGrid Statement of Charges 2025/26, SEMO, UREGNI, ESI/ECA Network Charges Report (Mar 2025)

Best Answer

With 24 months of empirical data, we can now be more precise. The median daily spread is EUR 61.7/MWh and round-trip costs are EUR 48–66/MWh, so the typical day is marginal (EUR ~0–15/MWh net). But the mean spread is EUR 75.7/MWh because high-spread days (P90 = EUR 143, P95 = EUR 170) pull the average up significantly — this is where most profit comes from. The top 25% of days generate 49% of total revenue.

Under forward scenarios, the median drops to EUR 28.7/MWh (combined 2028) or EUR 11.9/MWh (combined 2030), making pure arbitrage unviable without cost reform. Revenue stacking (DS3 system services, CRM capacity payments) remains essential for project viability under all scenarios.

Key Assumptions

• You can actually capture the full spread — in practice, forecast errors and market timing reduce capture rate
• DS3/FASS revenue continues near current levels — FASS competitive auction design is unknown, clearing price is unknown
• CRM de-rating factors don't change adversely

Why This Is Handwavy

The net spread calculation chains together a Reasonable spread estimate, a Solid fee schedule, and a Unknown capture rate. The output confidence is limited by the weakest link.

Sources: Calculated from above components

Best Answer

Grid-scale (50 MW / 200 MWh): EUR 58–150/kWh all-in, depending on source, configuration and whether turnkey or self-procured. EU turnkey systems currently EUR 91–108/kWh. Chinese FOB is dramatically cheaper but add 30–35% for shipping, duty, VAT.

Household (10 kWh): EUR 5,000–13,000 installed, depending on brand and configuration. No SEAI grant currently available for standalone battery.

Key Assumptions

• LFP cell prices (currently $36–56/kWh) stay flat or decline further — anti-dumping duties on BESS are being lobbied for
• Grid connection costs are site-specific and highly variable (EUR 1–10M+) — this is a placeholder range, not researched for specific sites Vibes

Sources: BloombergNEF, industry quotes, SEAI, Irish installer websites (Solar Energy Broker, RetrofitDublin)

Best Answer

• Wholesale arbitrage: EUR 60–72k/MW/year for a 4h system (Monte Carlo P10–P90, daily method); older estimate of EUR 31–60k/MW covered shorter durations and used profile-based capture assumptions
• DS3 system services: EUR 128–300k/MW/year historically — but this is declining. Oct 2024 scalar cuts reduced FFR revenue ~860 basis points. DS3 extended to 30 September 2027; DASSA replacement starts May 2027 at unknown prices.
• CRM capacity: EUR ~150k/MW/year (T-4 2028/29 clearing price confirmed)
• SDP wholesale access: "12–37% revenue uplift" per GridBeyond modelling — but NO post-launch empirical data exists yet. This is a projection, not a measurement. Handwavy
• DSO flexibility: Location-specific, 15-year contracts. ~109 MW first batch. Useful for stacking but not guaranteed.
• LDES procurement: EirGrid developing 201+ MW / 804+ MWh tender with floor-price + revenue share. CRU decision expected Q1 2026. If awarded, this is a game-changer for 4hr+ batteries. If delayed, it's nothing.

What Could Flip This

• FASS auction clearing prices could be much lower than DS3 administered rates — halving ancillary revenue
• CRM de-rating factors have been declining (2hr batteries now ~14% de-rated vs. higher historically)

Sources: SEM Committee Decision SEM-17-080, Everoze analysis, GridBeyond, SEMO T-4 auction results, EirGrid LDES consultation

Best Answer

Grid-scale: Our corrected model shows no payback under current policy (base-case IRR is negative at −3.5%; total undiscounted FCF = EUR −21.4M; MOIC 0.35x; NPV = EUR −22.7M at 5%). With D-TUoS reform from Year 3, IRR improves to ~8.7%; from Year 1, ~10.8%. Probability-weighted IRR across scenarios is ~1–2%. The previous model's "11.4% IRR" was inflated by ~5–6pp and did not reconcile with the cash flow table. D-TUoS charges alone are EUR 3,115k/yr (41.1% of gross revenue).

Household: 8–12 years on current TOU tariffs. Could compress to 5–8 years after dynamic pricing (June 2026) if automation captures spreads well. But this is a projection based on wholesale spreads being passed through — no one has tested it yet.

Why This Is Handwavy

Payback depends on: revenue (handwavy), costs (reasonable), fees (solid but may change), capture rate (unknown), degradation (reasonable but not field-tested in Ireland), and policy changes (unknown). The output is only as confident as the weakest input.

The "EUR 62/kWh/year" Residential Claim

This number appears in our household analysis. It's calculated from real tariff rates (solid) but assumes daily cycling, 90% efficiency, and 50–70% spread capture. The calculation is correct given those assumptions, but the assumptions themselves are optimistic for a typical household without sophisticated automation. Handwavy

Sources: Calculated from fee/cost/revenue research. See sensitivity analysis on profitability page.

Confirmed Timelines

• Dynamic tariff mandate: 1 June 2026. Five largest suppliers must offer half-hourly wholesale-linked pricing. Solid
• DS3 extended to: 30 September 2027 (was end-2026). Solid
• DASSA go-live: May 2027 (delayed from Dec 2026). Solid
• Celtic Interconnector: Spring 2028. 700 MW to France. Solid
• LDES procurement: CRU decision Q1 2026; contracts Q4 2027. Solid
• Carbon tax: EUR 71/tonne from May 2026; EUR 100/tonne by 2030. Solid

Uncertain But Important

• System Services charge reform: ESI/ECA proposed. CRU has not formally adopted. Timing and outcome unknown. Unknown
• FASS auction clearing prices: Design not yet published. Unknown
• Potential EU anti-dumping duties on BESS: Being lobbied for. No active investigation as of Feb 2026. Vibes

Sources: CRU Decision Papers, SEM Committee, EirGrid, Revenue.ie

Best Answer

There's a 10 GW pipeline but only 83 MW under construction. The pipeline is largely frozen because of DS3-to-FASS revenue uncertainty. Cornwall Insight forecasts 2.7 GWh (2025) growing to 13.5 GWh by 2030, implying ~5 GW discharge capacity.

We have now modelled specific saturation scenarios against 24 months of empirical spread data. The results show a clear trajectory: the combined 2028 scenario (Celtic Interconnector + incremental wind + 2 GW BESS fleet) compresses median spreads to EUR 28.7/MWh (54% decline from current EUR 61.7), with 96.2% of days still profitable. This is marginally viable for well-positioned projects. The combined 2030 scenario (all factors including full fleet buildout) drops median spreads to EUR 11.9/MWh (81% decline), with only 87.0% of days profitable and pure arbitrage revenue falling to ~EUR 27k/MW/yr — far below capex recovery requirements (vs ~EUR 111k/MW/yr currently).

Key Finding

The earlier claim that "pure arbitrage has a 2–3 year profitable runway" was speculation. The modelled numbers now support a more nuanced view: projects reaching COD by 2027–2028 face meaningful but survivable compression (median EUR 28.7/MWh, 96% profitable days), while projects relying on post-2030 arbitrage revenue face an unviable market for pure merchant plays. Revenue stacking remains essential in all scenarios. Handwavy — forward scenarios are modelled, not observed.

Celtic Interconnector Impact

We have now modelled Celtic's impact using peak-capping logic on the empirical dataset: median spread drops from EUR 61.7 to EUR 42.2/MWh (32% compression). This is still our own model based on peak-capping logic, not a published figure. The CRU/CEPA cost-benefit analysis says benefits "likely outweigh costs" but gives no EUR/MWh figure. Handwavy

Sources: Cornwall Insight, ESI Pipeline Survey, EirGrid, CRU/CEPA CBA (Celtic), 24-month spread analysis with forward scenario modelling