ERRATUM (2026-02-22): Financial model corrected. Prior versions stated base-case IRR of 6.6% and 11.4% — both were wrong. Corrected base-case IRR is negative (-3.5%). Total undiscounted FCF = EUR -21,356k (project never breaks even without reform). D-TUoS charges represent 41.1% of gross revenue. Recommendation changed from CONDITIONAL GO to DO NOT INVEST — WAIT.
Synthesis · The Verdict

Is This Investment Viable?

Corrected base-case IRR: -3.5% (no reform) → ~8.7% (with D-TUoS reform from Year 3), ~10.8% (reform from Year 1). The project is not viable without reform. D-TUoS charges consume 41.1% of gross revenue.

Do Not Invest — Wait
IRR (No Reform)
-3.5%
IRR (Reform Yr 3)
8.7%
NPV @5%
-22.7M
Simple Payback
Never
MOIC
0.35x
IRR (Reform Yr 1)
10.8%
How was this page built?
1
20-Year DCF Model
Standard project finance discounted cash flow model. Year 1 revenue from the Revenue Model page, escalated at 2% inflation. CAPEX from the Hardware page. OPEX from the Fee Structure page. Discount rate: 8% (typical BESS WACC).
2
Two-Regime Comparison
Ran the model twice: once with current D-TUoS charges (EUR 3,115k/yr, 41.1% of gross revenue) and once assuming reform. Without reform, IRR is -3.5%. With reform from Year 3: ~8.7%; from Year 1: ~10.8%.
3
Cash Flow Waterfall
Year-by-year: Gross Revenue EUR 9,434k – D-TUoS EUR 3,115k (41.1%) – other OPEX = negative net cash flow under current regime. Total undiscounted FCF: EUR -21,356k. Project never breaks even without reform.
4
Verdict Matrix
Combined all upstream inputs into a 3×2 decision matrix (3 scenarios × 2 D-TUoS regimes) to show which combinations produce viable projects (IRR > 8%).
Limitations: The 20-year horizon assumes battery degradation is managed through augmentation (replacement of ~20% of cells at year 10). Revenue escalation at 2% is optimistic if spreads compress. The model is deterministic — it doesn’t capture revenue variance (see the Price Modeling page for the probabilistic view). D-TUoS reform timing is the single largest source of uncertainty.
Nerd level: DCF models are bread-and-butter finance — the hard work was getting the inputs right
Show figures as
01 The Verdict Matrix 3 scenarios × 2 regulatory regimes

Every cell shows the unlevered IRR and NPV at 8% WACC. The project requires 8% IRR to create value. Colour coding: red = NO-GO, yellow = CONDITIONAL, green = GO. C3

Current D-TUoS Regime
With D-TUoS Reform
Bear
No-Go
N/A
NPV: EUR -36.5M
No-Go
-2.2%
NPV: EUR -19.1M
Central
No-Go
-3.5%
NPV @5%: EUR -22.7M
Conditional (Yr 3)
8.7%
Reform from Yr 1: 10.8%
Bull
Go
18.5%
NPV: EUR +30.7M
Strong Go
24.2%
NPV: EUR +48.2M
C1 D-TUoS rates: published EirGrid tariffs C2 CRM auction: SEM Committee result C3 Capture rate, de-rating: GB proxy C4 DASSA, cycling rate: estimates

IRR/NPV: scripts/compute_irr.py. Capex EUR 34.0M [CAPEX-BUILD-UP.md]. 20-year life, no terminal value, no tax, 8% unlevered WACC. Full data: PROFITABILITY-SYNTHESIS.md §3.2.

02 Revenue vs Cost (Year 1, 2028) Central scenario, EUR thousands
Year 1 Revenue and Opex by Regime Central scenario, EUR k
Capex recovery requirement: EUR 3,463k/yr at 8% WACC over 20 years (annuity of EUR 34,000k capex). D-TUoS charges of EUR 3,115k/yr represent 41.1% of Year 1 gross revenue (EUR 9,434k). Total undiscounted FCF over 20 years: EUR -21,356k — the project never breaks even under the current regime. MOIC is 0.35x, confirming negative IRR (-3.5%). To achieve 8% hurdle without reform, spreads must exceed EUR 90/MWh.
C1 D-TUoS rates: EirGrid Statement of Charges 2025/26 C2 CRM: T-4 2028/29 auction, EUR 149,960/MW de-rated C3 Arbitrage: EUR 86/MWh gross × 65% capture C4 DASSA: EUR 20k/MW/yrEUR 1.0M/yr, no auction held

Revenue: REVENUE-BUILD-UP.md §2. Opex: ANNUAL-COSTS.md §5.1-5.2. D-TUoS: FEE-VERIFICATION.md §3.1. Capex annuity: EUR 34M × CRF(8%, 20yr) = EUR 3,463k/yr.

03 20-Year Cumulative Cash Flow Central scenario, undiscounted, EUR thousands
Cumulative Cash Flow Current regime vs Reform — Central scenario

Corrected: Under the current regime, the project never breaks even. End-of-life cumulative FCF: EUR -21,356k (current regime). With D-TUoS reform from Year 1, payback occurs around Year 7. The project is ONLY viable with reform. Total undiscounted FCF (current) = EUR -21,356k; MOIC = 0.35x.

C1 Opex rates: published tariffs C3 Revenue: spread projection + GB capture proxy C4 Cycling rate: no 4hr data globally

Cash flows: PROFITABILITY-SYNTHESIS.md §3.3-3.4. Reform saving: FEE-VERIFICATION.md §6.2.

04 Where The Money Goes Annual breakdown, Year 1 (2028), Central / Current regime
Revenue vs Cost Stack Year 1, EUR k
Arbitrage
CRM
DASSA
D-TUoS Sys. Svc.
D-TUoS Capacity
D-TUoS Transfer
Other Costs
D-TUoS System Services alone costs EUR 1.97M — more than all of DASSA revenue (EUR 1.0M) plus half of O&M (EUR 0.21M). This single charge line consumes 26% of total revenue under the current regime. It is the reason reform is so critical.
C1 D-TUoS: EirGrid Statement of Charges 2025/26 C2 CRM: T-4 2028/29 auction result C3 Arb: EUR 86/MWh gross × 65% capture × 62,050 MWh C4 DASSA: EUR 20k/MW/yrEUR 1.0M/yr — no auction held

Revenue: REVENUE-BUILD-UP.md. Costs: ANNUAL-COSTS.md §5.1. D-TUoS: FEE-VERIFICATION.md §3.1. CRM: c004-crm-auction-results.md.

05 Evidence Chain Every number traces back to its source

Each revenue and cost line is decomposed into constituent assumptions. Follow the arrows leftward to reach primary data. Confidence tags indicate the solidity of each link in the chain.

Arbitrage EUR 3.45M/yr C3
EUR 3,452k ← EUR 86/MWh gross spread C2 × 65% capture rate C3 × 62,050 MWh/yr ← 0.85 cycles/day C4 × 200 MWh × 365 days
Gross spread: EUR 76/MWh historic baseline [SPREAD-PROJECTION.md §1.4; 64,670 hourly SEM prices, Jan 2024 – Jan 2026] + EUR 9.6 net adjustment (Celtic -3.0, BESS +0.0, RES-E +15.6, Gas -3.0) [SPREAD-PROJECTION.md §3.2, Appendix A]. Capture rate: 65% = GB 2hr operational benchmark, Modo Energy 2024 [KEY-FINDINGS.md §Q1; GB-BESS-PERFORMANCE.md §2.2]. This represents an average professional operator. ML-optimised trading (75–85% of PF) would push arbitrage to EUR 79–96k/MW/yrEUR 3.95–4.8M/yr (+EUR 500k–1.35M/yr+EUR 0.5–1.35M/yr), swinging IRR by +1.5–3.5pp. Cycling: 0.85/day = GB 2hr proxy (1.0/day measured) minus duration penalty; no 4hr data exists globally [KEY-FINDINGS.md §Q2]. ML dispatch could also increase effective cycling to 0.95–1.05/day by identifying profitable second cycles.
CRM (Capacity) EUR 3.0M/yr C2
EUR 2,999k ← EUR 150k/MW de-ratedEUR 3.0M total C1 × 20 MW de-rated ← 50 MW installed × 40% de-rating C3
Clearing price: T-4 2028/29 CRM auction, EUR 149,960/MW de-rated [c004-crm-auction-results.md; SEM Committee published, C1]. De-rating: 40% estimate for 4hr BESS, SEM Committee methodology post-halving [c004; GB-BESS-PERFORMANCE.md §6.3, C3]. Exact factor requires unpublished IAIP methodology document.
DASSA (System Services) EUR 1.0M/yr C4 C5
EUR 1,000k ← EUR 20k/MW/yrEUR 1.0M/yr C4 × 50 MW ← GB proxy EUR 12-19k/MW/yrEUR 0.6–0.95M/yr + Ireland premium
CRITICAL GAP: No DASSA auction has been held. No clearing price exists. First auction expected May 2027 [KEY-FINDINGS.md §Data Gaps]. GB ancillary: ~20-25% of total BESS revenue, ~GBP 10-16k/MW/yr = EUR 12-19k/MW/yrEUR 0.6–0.95M/yr [Modo Energy 2024, GB-BESS-PERFORMANCE.md §1.2]. Range: EUR 10-40k/MW/yrEUR 0.5–2.0M/yr. Central EUR 20kEUR 1.0M includes Ireland premium for higher RES-E volatility.
D-TUoS (Network Charges) EUR 3.48M/yr C1
EUR 3,476k total ← Sys. Svc. EUR 1,971k + Capacity EUR 1,206k + Transfer EUR 299k ← Published EirGrid rates C1 × 73,000 MWh imported
Rates: EirGrid Statement of Charges 2025/26 (C1): System Services EUR 26.47-29.12/MWh (80/20 peak blend = EUR 27.00), Network Transfer EUR 4.09/MWh, Capacity EUR 2,009.70/MW/month. Volume: 200 MWh/cycle / 0.85 RTE × 0.85 cycles/day × 365 = 73,000 MWh [ANNUAL-COSTS.md; FEE-VERIFICATION.md §3-4]. Market fees: EUR 64k (SEMO EUR 0.61/MWh, C1). O&M: EUR 425k. Insurance: EUR 170k [ANNUAL-COSTS.md §2.2-2.3].
06 The Three Things That Must Be True For the investment to work
1

"D-TUoS reform happens"

+12.3pp IRR impact (from -3.5% to ~8.7% with Yr 3 reform)
Probability by Oct 2027: 20-30%

The single largest variable. Without reform, central IRR is -3.5% (negative — NO-GO). D-TUoS charges are EUR 3,115k/yr = 41.1% of gross revenue. With reform from Year 3: ~8.7% IRR. With reform from Year 1: ~10.8% IRR. The CRU has not published a Phase 2 consultation as of Feb 2026. ESI/ECA recommended System Services exemption in Mar 2025. Every year of delay costs EUR 1.5-2.0M in foregone savings. The project is ONLY viable with reform.

Source: CRU consultation timeline, FEE-VERIFICATION.md §7.3 C2
2

"Spreads don't collapse"

EUR 84/MWh needed vs EUR 76/MWh historic
BESS cannibalisation is the primary threat

Central case requires EUR 86/MWh gross spread (2028). Historic average is EUR 76/MWh. The EUR 9.6/MWh gap is filled by RES-E widening (+15.6) minus Celtic IC (-3.0) and gas (-3.0). If BESS fleet exceeds 3 GW by 2030, cannibalisation could overwhelm these tailwinds. Cannibalisation slope: -EUR 11/MWh per GW deployed [SPREAD-DRIVERS.md].

Source: SPREAD-PROJECTION.md §3.2-3.3, SPREAD-DRIVERS.md C3
3

"DASSA delivers EUR 20k+/MW/yrEUR 1.0M+/yr"

Range: EUR 10-40k/MW/yrEUR 0.5–2.0M/yr
Status: No auction held; pure estimate

DASSA replaces the DS3 regulated tariff system. At EUR 10k/MW/yrEUR 0.5M/yr (bear), IRR drops 2.1pp. At EUR 40k/MW/yrEUR 2.0M/yr (bull), the project clears the hurdle rate even without D-TUoS reform (+3.9pp). First auction expected May 2027. GB proxy suggests EUR 12-19k/MW/yrEUR 0.6–0.95M/yr before Ireland premium.

Source: KEY-FINDINGS.md §Q3, GB-BESS-PERFORMANCE.md §1.2 C4 C5
07 Comparison to Previous Model Why IRR dropped from 11.4% to -3.5%
Assumption-by-Assumption Comparison
Assumption Old Model New Model Impact Source / Conf.
IRR (current regime) 11.4% -3.5% -14.9pp scripts/compute_irr.py C3
Arbitrage/MW/yr/yr EUR 80,000/MWEUR 4.0M EUR 69,049/MWEUR 3.45M -14% REVENUE-BUILD-UP.md C3
Cycles/day 1.25 0.85 -32% KEY-FINDINGS.md §Q2 C4
D-TUoS fees/yr EUR 3.1M EUR 3.54M +14% FEE-VERIFICATION.md C1
D-TUoS reform assumed Year 2 Not assumed Major FEE-VERIFICATION.md §7.3 C2
DASSA/MW/yr/yr EUR 30,000/MWEUR 1.5M EUR 20,000/MWEUR 1.0M -33% KEY-FINDINGS.md §Q3 C4
The old model overstated IRR by ~15pp through compounding errors: optimistic cycling (1.25 vs 0.85 cycles/day), understated D-TUoS fees (applied to export not import volume, and assumed early reform), and an aggressive DASSA assumption (EUR 30k vs EUR 20k/MW/yrEUR 1.5M vs EUR 1.0M/yr). The old model was not fabricated — its implied ~EUR 85/MWh gross spread matches our data — but it was an extreme upside scenario presented as a base case. The corrected base-case IRR is -3.5% (negative), with MOIC of 0.35x and total undiscounted FCF of EUR -21,356k. The project never breaks even without reform. The old model's 11.4% IRR is recoverable only with D-TUoS reform from Year 1 (~10.8%), which requires the regulatory change that has not occurred. Recommendation: DO NOT INVEST — WAIT for reform certainty.
C2 Old model: REVIEW-AND-REVISED-ANALYSIS.md C3 New model: scripts/compute_irr.py

Old model: REVIEW-AND-REVISED-ANALYSIS.md lines 29-31, 204-217. New model: PROFITABILITY-SYNTHESIS.md §8. IRR gap reconciliation: §8.2.

Analysis basis: 50 MW / 200 MWh LFP BESS, 4-hour duration, 20-year project life (2028-2047). Capex EUR 34.0M [CAPEX-BUILD-UP.md]. WACC 8% unlevered. No tax, no terminal value, no leverage. All numbers trace to source documents referenced in [brackets]. IRR and NPV computed by scripts/compute_irr.py with full 20-year cash flows. Confidence scale: C1 (published official data) through C5 (guess/no data).

Data sources: 7 years of SEM hourly price data (64,670 records), GB operational benchmarks (Modo Energy 2024), EirGrid/SEMO published tariff schedules (C1), SEM Committee CRM auction results, independent capex build-up from BNEF/InfoLink/Ember.

Prepared: 2026-02-20. For investment decisions, obtain independent bankable revenue forecasts from Cornwall Insight or LCP Delta, and project-specific cost quotations from EPC contractors. This analysis is for research purposes and does not constitute investment advice.