Fact check · claim about Irish data centres
partly true (size uncertain)"Data centres are making my electricity bill higher."
Verdict: partly true (size uncertain). Probably yes through wholesale prices, by an amount that depends on the model. The Fearon/FoE study put it at ~EUR 360 per household over 2015-23; our rerun of its method on independent data gives a similar ~EUR 290 for 2021-23 alone and ~EUR 110-130 a year now, but the method holds imports and power plants fixed, so treat it as an upper-end estimate (CRU2025198, Dec 2025).
Checked against data built 10 October 2026 · how we check claims · corrections
The full explanation
Probably yes through wholesale prices, by an amount that depends on the model. The Fearon/FoE study put it at ~EUR 360 per household over 2015-23; our rerun of its method on independent data gives a similar ~EUR 290 for 2021-23 alone and ~EUR 110-130 a year now, but the method holds imports and power plants fixed, so treat it as an upper-end estimate. The PSO levy is split by share of peak demand, so households pay 34% of it while using 27.5% of electricity, roughly EUR 5 a year each. Network charges cut both ways: extra-large users' rose 175% from 2020-21 to 2024-25 (households 31%), but from 2024-25 to 2029-30 they are set to fall 20% while households' rise 15%.
Sources
| Figure | Value | Source | Date |
|---|---|---|---|
| CRU archetypes: domestic vs Extra Large Energy User (190 GWh/yr, 44 MW - data-centre sized). The base year flips the story. | domestic_2024_25_to_2029_30_pct: 15 · xleu_2024_25_to_2029_30_pct: -20 · domestic_2020_21_to_2029_30_pct: 51 · xleu_2020_21_to_2029_30_pct: 119 | CRU2025198, PR6 Final Determination - Impact Analysis Note, Tables 1 and 16 | 2025-12-16 |
Also: data/raw/press/sources/fearon-replication-2026-10-03.json · data/raw/press/sources/cru-CRU2025108-PSO-2025-26-final-2026-10-03.txt · data/raw/press/reports.json (fact-checks 3 Oct 2026)
Press claims on this, fact-checked
| Claim | Rating | Why |
|---|---|---|
| Large energy users' network tariffs rose much faster than households': +175% vs +45%, 2020-2026. Irish Examiner, 2026-09-29 | mostly accurate | The +175% reproduces exactly from the CRU's PR6 impact note: the extra-large energy user archetype (190 GWh/yr, 44 MW, data-centre sized) paid EUR 2.45m in network charges in 2020-21 and EUR 6.76m in 2024-25 (+175%). The same tables give the domestic archetype EUR 293 -> EUR 384 (+31%), not +45%. We could not reproduce +45% (it may use 2025-26 tariffs; unchecked). |
| CRU rules socialise data-centre grid costs onto households; connection fees don't cover reinforcement (TCD, 'By Design, Not by Default'). Irish Times, 2026-09-22 | can't be tested with public data | No public data splits network reinforcement cost by the customers who caused it, so the core claim cannot be tested. Supporting facts check out. CRU (CRU202504, Feb 2025): Castlebaggot 220 kV station was planned for housing and West Dublin growth, but its new capacity was 'almost entirely used by data centres'. The EUR 100m cost was not checked. CRU's 2020 factsheet put network costs at 31.8% of a typical household bill ('~30%'). |
| Data-centre demand pushed up wholesale prices: ~EUR 360 per household over 2015-2023. RTÉ, 2026-05-28 | mostly accurate | Mechanism: data centres add a flat ~0.9 GW to demand, so more hours are priced by gas plant. Our wholesale data agree gas most likely set the Irish price in ~81% of half-hours (Oct 2025-Sep 2026). We reran the same residual-demand method on independent data. 2021-23 (Ember hourly prices + EirGrid demand and wind): ~EUR 293 per household (3,682 kWh), vs Fearon's EUR 263. 2020: ~EUR 23. Oct 2025-Sep 2026 (SEMOpx + EirGrid): ~EUR 30/MWh, or ~EUR 110-130 per household per year. |
| Data centres added ~EUR 360 to the average household's electricity costs over 2015-2023 (EUR 263 in 2021-23) and will add EUR 295-644 over 2025-2034. Friends of the Earth Ireland, 2026-05-27 | mostly accurate | Mechanism: data centres add a flat ~0.9 GW to demand, so more hours are priced by gas plant. Our wholesale data agree gas most likely set the Irish price in ~81% of half-hours (Oct 2025-Sep 2026). We reran the same residual-demand method on independent data. 2021-23 (Ember hourly prices + EirGrid demand and wind): ~EUR 293 per household (3,682 kWh), vs Fearon's EUR 263. 2020: ~EUR 23. Oct 2025-Sep 2026 (SEMOpx + EirGrid): ~EUR 30/MWh, or ~EUR 110-130 per household per year. |
| Data-centre demand raises power-system costs by 6% (carbon policy) to 9-15% (renewable targets). ESRI, 2020-04-02 | can't be tested with public data | This is a counterfactual optimisation result; no public dataset can reproduce it. One premise has changed since 2020: the extra demand was not met mainly by new domestic plant. From 2020 to 2025 data centres used 4.6 TWh more, wind was flat at ~11.8 TWh, Irish gas generation fell, and net imports rose from ~0 to ~6.1 TWh (Ember; EirGrid's ROI series gives -0.5 to 5.8 TWh). |
| The PSO levy is allocated by contribution to peak demand, so households and SMEs pick up part of the extra renewable subsidy that data-centre demand creates. ESRI, 2019-06 | mostly accurate | The mechanism still holds. CRU's 2025/26 PSO decision allocates the levy across domestic, small and medium/large customers 'in proportion to the ratio of these demand peaks'. Households carry 34.23% of the levy but used 27.5% of metered electricity in 2025 (CSO MEC05). Medium/large users, including data centres, carry 55.71% while using ~62%. Flat loads like data centres add little to the peak, so they pay less than their energy share. |
Each claim is tested as stated against public data; ratings: accurate, mostly accurate, misleading, wrong, or can't be tested with public data. Everything reported