Ireland Electricity System

Electricity Demand Analysis

Ireland's demand is entering a period of structural growth unlike anything in its modern history. Data centres dominate, electrification accelerates, and peak demand is breaking records.

Demand growth accelerating beyond forecasts
Total demand (2024)
31.9 TWh
Growth from DCs
88% of 2015-2024
Forecast increase
+45% by 2034
Record peak (Jan 2025)
6,024 MW
01 Historical Demand Trajectory 2015 -- 2024, CSO metered consumption

Total Demand: Data Centre vs Non-DC C1

C1: CSO metered data (gold standard)

Non-data-centre metered demand has been essentially flat at ~24,000-25,000 GWh for a decade. Virtually all growth is attributable to data centres, which grew from 5% to 22% of metered electricity between 2015 and 2024 -- a 463% increase in absolute consumption.

Metric Definitions

Confusion often arises from comparing figures that measure different things:

Metric 2024 Value What It Captures
CSO Metered Consumption 31.9 TWh Customer meter readings only
SEAI End-User Demand 32.9 TWh Metered + adjustments for unmetered loads
Gross Electricity Supply ~36 TWh Includes losses (~8-9%), own-use, imports
Peak System Demand 6,024 MW Instantaneous maximum

The ~4 TWh gap between gross supply and metered consumption represents transmission/distribution losses and generator own-use. EirGrid's "45% growth" refers to Total Electricity Requirement (including losses).

02 Peak Demand Records & Temperature Sensitivity Ireland's system stress points

Peak Demand Record Timeline C1

Date Peak (MW) Notes
Nov 2016 ~4,760 21 November 2016
Dec 2022 5,544 Previous multi-year record
Jan 2024 5,577 18 January 2024
Nov 2024 5,639 20 November 2024
Jan 2025 6,024 8 January 2025 -- first time above 6 GW Record
+40-55 MW
Per 1°C drop in outside temperature
Temperature sensitivity of peak demand
17:00-18:00
Winter evening peak window
Convergence of residential, commercial, and heating loads
>7,000 MW
Projected peak by 2034
EirGrid AIRAA median scenario
40-50%
Summer minimum vs winter peak
Seasonal demand swing creates arbitrage opportunity

How Each Driver Affects Peak vs Baseload C2

Driver Load Profile Peak Impact Baseload Impact
Data Centres Flat 24/7 Moderate Raises floor High Constant draw
Heat Pumps Winter-weighted, cold-correlated Very High Aligned with peak Moderate Winter only
EVs (unmanaged) Evening spike High 5-8pm coincidence Low
EVs (smart charged) Overnight shift Low Moderate Fills overnight valley
Housing Growth Residential profile Moderate Moderate
Industrial Daytime weekday Moderate Shifts to 1-3pm High
Critical risk: The combination of heat pumps and EVs during a cold winter evening. At full CAP deployment (680k heat pumps + 945k EVs), unmanaged coincident load could add 2,000-3,000 MW to peak demand. Even with diversity and smart management, the addition could be 800-1,500 MW by 2030.
03 Demand Component Deep Dive Six drivers of Ireland's demand growth

Data Centres -- The Dominant Driver C1

22%

Data centres grew from 5% to 22% of metered electricity (2015-2024), a 463% increase in absolute consumption (1.2 to 7.0 TWh). Ireland hosts ~121 operational data centres; Dublin has Europe's second-largest cluster (~1,150 MW in operation). 97% are in the Greater Dublin Area.

5.8 GW
Expression-of-interest pipeline
Nearly equal to entire record peak demand (6,024 MW). Realistic realisation: 30-60% (1.7-3.5 GW)
10%
2024 DC demand growth (decelerating)
Down from 20-32% in prior years due to Dublin moratorium
14.6 TWh
Projected DC demand by 2034
31% of total demand (EirGrid AIRAA median)
97%
DCs in Greater Dublin Area
Extreme concentration vs 44% in Amsterdam

CRU LEU Connection Policy (Dec 2025) C1

The CRU's decision paper CRU/2025/236 ended the 4-year moratorium and established a tiered framework requiring self-generation and 80% renewable sourcing for new data centre connections.

TierMIC ThresholdSelf-GenerationRenewables
1 (De minimis)< 1 MVAExemptExempt
2 (Medium)1-10 MVAAutoproducer at 100% MIC; SEM participation80% new Irish renewables
3 (Large)≥ 10 MVAOnsite dispatchable gen/storage matching MIC; SEM80% new Irish renewables

Self-generation will reduce grid draw during stress events but will not reduce total energy consumption. Most compliance will use on-site gas turbines: Microsoft is building 170 MW at Grange Castle; Digital Realty invested EUR 100m in a 9 MW gas plant. The 80% renewable obligation should drive 2-4 GW of new wind/solar via CPPAs.

Electric Vehicles C2

~1.5%

Ireland reached 200,000 EVs by end-2025 (BEVs outsold diesel for the first time). Per-vehicle consumption: ~2,500-3,000 kWh/year. ~80% charged at home.

MetricValue
CAP 2030 target945,000 EVs
Current fleet (2025)~200,000
Gap~745,000
Required annual adds~185,000/yr
Best actual year23,601 BEVs (2025)
Acceleration needed~8x current

Realistic 2030: 350,000-500,000 EVs (40-55% of target). Demand: ~1.0-1.2 TWh. Unmanaged peak impact: up to 500 MW; smart charging reduces to 50-100 MW. Only 11% of customers on smart tariffs. C3

Heat Pumps C2

~1.5-3%

CAP target: 680,000 by 2030. Current: ~150,000-180,000. New builds on track (96% electric heating). Retrofits critically behind (need ~20x acceleration).

Segment2025 TargetAchieved2030 Target
New dwellings170,000~153,000280,000
Existing (retrofit)45,000~15-17k400,000
Total215,000~150-180k680,000

Winter peak concern: 40-45% of annual HP consumption in Dec-Feb. After-diversity max demand per HP: ~2.3 kW (typical), rising to ~3.8 kW in extreme cold (+65%). At 680k HPs, peak addition: 1,565-2,585 MW. C3

Feb 2026 grant increase (up to EUR 12,500, 80% of costs) is most significant intervention yet.

Industrial & Commercial C2

~21%

Manufacturing accounts for ~6.7-7.0 TWh (~21% of total). Key consumers:

  • Intel (Leixlip): ~130-140 MW average load. Fab 34 (EUR 17bn) could increase use 50%.
  • Pharma: 75 companies, 99+ facilities. Pfizer EUR 1.2bn (Grange Castle, 2027); Eli Lilly USD 2bn (Limerick, 2026).
  • Aughinish Alumina: ~40 MW. Suspended from energy market July 2025 (Russian ownership).
  • SEAI LIEN: 200+ companies, EUR 2.2bn annual energy spend.

CAP targets 3.5 TWh of new industrial heat pumps (would add ~10-11% to total demand). LEAP introduces Green Energy Parks co-locating industry with renewables at "hundreds of MW" scale.

Residential C1

25.5%

Total residential metered: 8,951 GWh (2024). Median per-household: 3,246 kWh/year. 10-year growth just 1.7%. ~2.24 million meters.

  • Housing growth: Target 303,000 new homes by 2030 (60,000/yr). Actual 2025: ~36,000. Each NZEB home: ~5,000 kWh/yr.
  • Microgeneration: 155,000+ rooftop solar installations (800+ MW). Net metered reduction ~200-350 GWh.
  • Smart meters: 1.9M installed (83% coverage), but only 11% on smart tariffs. 58% unaware they need to switch.
  • Energy poverty: 300,000 homes in arrears. Avg bill nearly doubled in 5 years (EUR 976 to EUR 1,817).

EirGrid projects residential to grow ~7 TWh to 2035 (from ~9 to ~16 TWh), driven by HPs, EVs (home charging), and housing. Second-fastest sector after DCs. C2

Hydrogen C2

<0.1%

Not material through 2030. No large-scale electrolysers operational. Total firm pipeline: ~130-135 MW, none at FID. Ireland lacks heavy industrial anchor customers.

PeriodRealistic DemandGrid Impact
2026-20300-0.5 TWhNegligible
2030-20350.5-4 TWhNoticeable
2035-20404-15 TWhSignificant
2040-205015-40 TWhTransformational C4

For modelling through 2034, hydrogen is a sensitivity/upside scenario, not base case. Longer-term: Arup estimates 215-430 kt/yr from 3.75-7.5 GW offshore wind (7-24 TWh). C3

04 Demand Growth by Driver Incremental TWh, 2024 to 2034 (median)

Share of Growth C3

Total incremental growth: ~14.9 TWh from 2024 to 2034 (median scenario). Data centres account for over half of all growth despite being only 22% of current demand.

Demand Composition -- Median Scenario C3

Component 2024 2027 2030 2034
Data centres 7.0 ~9.5 ~12.0 ~14.6
Residential (incl. HPs, EVs) 8.9 ~9.5 ~11.0 ~13.0
Industry 6.8 ~7.0 ~7.5 ~8.0
Commercial (non-DC) 6.5 ~6.5 ~6.5 ~6.5
Transport (EVs, public) 0.4 ~0.7 ~1.2 ~2.5
Other (agri, public) 3.3 ~3.3 ~3.3 ~3.2
Total 32.9 ~36.5 ~41.5 ~47.8

Indicative allocations. EirGrid's full sectoral breakdown is not publicly available at this granularity. These are synthesis estimates from d01-d07 with EirGrid AIRAA shares applied.

05 Demand Scenarios: Low / Median / High 2024 -- 2034, total electricity requirement (TWh)
2024: SEAI provisional actual 2025-2027: Near-term estimates (C2) 2028-2034: Scenario projections (C3)

All forecasters agree demand grows substantially; no scenario shows flat or declining demand. The realistic range for 2030 is 37-45 TWh; for 2034, 42-50 TWh. The IEA's pathway suggests demand could "nearly double" by 2035 (most aggressive estimate).

06 Source Agreements & Disagreements Where EirGrid, SEAI, CSO, and IEA diverge

Where Sources Agree C1

Topic Consensus
Direction of growth All forecasters agree demand grows substantially; no scenario shows flat or declining demand
DCs as dominant driver Universal agreement that DCs drove most historical growth and will continue to lead
2024 DC share (22%) CSO gold-standard measurement; accepted by all
Peak demand records EirGrid real-time data; undisputed
CAP targets will be missed EPA, CCAC, multiple analysts agree Ireland is off track for 2030 EV, heat pump, and renewable targets
Grid adequacy 2025-2027 EirGrid, CRU, and CCAC all flag this period as "potentially challenging"

Where Sources Disagree C3

Topic EirGrid SEAI/CSO IEA
DC share of demand 2030 ~23% (AIRAA) ~30% (SEAI blog) 32% by 2026
Total demand 2030 41 TWh (median) Not directly forecast Higher (doubling pathway)
Residential growth by 2035 +7 TWh Not quantified --
Demand vs GDP Decoupled (structural) Historical correlation --
The DC share disagreement (23% vs 30% vs 32%) is significant. EirGrid's 23% for 2030 implies total demand grows fast enough to dilute the DC share even as DC absolute demand grows. IEA's 32% assumes slower total growth with faster DC connections. The actual outcome depends primarily on how quickly new data centres connect post-moratorium.
07 Policy Decision Dependencies Decisions that will materially shape demand trajectory
Policy Demand Impact Status
CRU LEU Implementation Determines pace of new DC connections; each 100 MW DC = ~0.7-0.8 TWh/year. System operator processes due 31 March 2026. Published; pending
Dublin Moratorium End Unlocks largest cluster; could add 500+ MW of DC load. Expected "at least 2028". Grid-dependent
EV Purchase Grants & BIK 2024 BIK uncertainty caused 22% sales drop. Currently EUR 3,500 grant; VRT relief to EUR 5,000. Annual budget
Heat Pump Grants Feb 2026: up to EUR 12,500 (80%). Could triple retrofit rates if effective. Just announced
Gas Boiler Ban EU EPBD requires phase-out by 2040. Transposition timeline critical for retrofit acceleration. No Irish ban yet
Housing Delivery Each 10,000 homes = ~50 GWh/year. Target 60,000/yr by 2030; actual 2025: ~36,000. Behind target
Smart Tariff Adoption Key to managing EV/HP peak impact. Only 11% uptake despite 83% smart meter coverage. 11% uptake
LEAP Green Energy Parks Enables large industrial/DC load outside Dublin. "Hundreds of MW" scale. Entering planning
Carbon Price Trajectory Affects competitiveness of electrification vs fossil. Currently ~EUR 60-70/tonne; hydrogen needs EUR 150-200. EU ETS driven
08 BESS Arbitrage Implications How demand trends affect battery storage profitability

Demand-Side Drivers of Battery Profitability C4

Widening peak-trough spreads: Heat pumps and EVs increase winter evening peak while data centres raise overnight baseload. Smart EV charging shifts load overnight, making evening peak more pronounced. Net effect: winter evening peak grows faster than overnight demand, widening daily price spreads.
Temperature-driven demand spikes: Heat pump correlation with cold weather creates sharper, weather-dependent peak events. Cold anticyclonic conditions mean high demand + low wind -- price spikes that BESS can arbitrage.
DC self-generation creates new dynamics: On-site gas generation participating in SEM creates more arbitrage windows. Surplus wind absorbed by DCs at low prices, followed by peaks when DC self-gen is dispatched at marginal gas prices.
Rising renewable curtailment: Projected 11-29% curtailment creates periods of very low/negative prices ideal for BESS charging. Growing demand does not fully resolve curtailment due to SNSP limits and minimum generation constraints.
DSM gap creates vacuum: Target: 20-30% flexible demand by 2030. Reality: ~600 MW from DRAI members. Until demand-side flexibility scales, BESS is the most deployable flexibility source. EirGrid warns of "mandatory demand curtailment" without more flexibility.
Adequacy gaps from 2028-2032: EirGrid flags 600-800 MW adequacy shortfalls. Capacity payments and reliability-driven dispatch will be increasingly valuable for BESS beyond energy arbitrage.

Risks to BESS from Demand Trends C3

Smart charging competition: If V1G/V2G succeeds at scale, EV batteries compete directly with grid-scale BESS for peak-shaving revenue.
DC self-gen caps peak prices: Data centre on-site generation adds supply during stress events, potentially capping the very peak prices that drive arbitrage.
Low-scenario demand growth: If demand reaches only 37 TWh by 2030 (low scenario), reduced price volatility dampens arbitrage returns.
Hydrogen as flexible demand: By the 2030s, electrolysers operating flexibly could absorb low-price periods that BESS relies on for cheap charging.
Net assessment: Structural demand growth is positive for BESS arbitrage in the medium term (2025-2030). The system needs flexibility that batteries provide. However, magnitude of opportunity is uncertain: it depends on which shortfalls materialise, how fast BESS deploys (currently stalled at 83 MW under construction), and whether market design evolves to remunerate flexibility properly. First-mover advantage is real but time-limited.
09 Confidence Ratings Summary C1 (highest) to C5 (speculative)

Confidence Scale

C1 Direct measurement / official data
C2 Well-supported estimates
C3 Inference / interpolation
C4 Directional inference, high uncertainty
C5 Speculative
Claim Rating Justification
Historical demand data (2015-2024) C1 CSO metered data -- "gold standard" (SEAI's description). Directly measured from ESB Networks meters.
DC share of demand (22%, 2024) C1 CSO direct measurement; corroborated by all sources.
Peak demand records (6,024 MW) C1 EirGrid real-time system data.
CRU LEU policy framework C1 Published decision paper with multiple legal analyses confirming.
EirGrid median: ~41 TWh by 2030 C2 Well-constructed scenarios but wide range (37-45 TWh). DC-timing-dependent.
945k EV target will be missed C2 EPA, CCAC, Irish Times, and multiple analysts agree. Trajectory implies 350-500k by 2030.
680k heat pump target will be missed C2 Retrofit stream critically behind (~20x acceleration needed). EPA projects 571k.
Hydrogen negligible through 2030 C2 No large electrolysers under construction; pipeline ~130 MW, none at FID.
Year-by-year trajectory 2025-2034 C3 Interpolated from anchor points; depends on lumpy DC connections.
DC share disagreement (23%/30%/32%) C3 Different methodologies; resolution depends on connection timing.
5.8 GW pipeline realisation (30-60%) C3 Inference from historical conversion rates + new CRU requirements.
HP + EV peak impact (800-1,500 MW) C3 Calculated from per-unit ADMD data with diversity assumptions; not validated at scale.
BESS arbitrage implications C4 Directional inferences from demand trends; actual profitability depends on supply-side factors and market design.
2050 demand of ~75 TWh (MaREI) C4 Academic modelling of Net Zero pathway; 25-year horizon with enormous uncertainty.
Hydrogen export economy (late 2030s) C4 Government ambition backed by resource assessment but no infrastructure, no FID, no confirmed demand.

Data Sources

This analysis is compiled from 8 research files (d01-d08) drawing on official statistics, government forecasts, and published analyst reports. All claims are sourced from input files; no web searches were performed during synthesis. Confidence ratings follow the C1-C5 scale where C1 = highest confidence (direct measurement/official data), C5 = speculative. Projections are estimates; actual outcomes will differ. Ireland's GDP data is distorted by multinational activities; Modified Domestic Demand (MDD) is a better comparator for economic-demand correlation.