Electricity distribution scorecard
A quick look at how New Zealand's electricity lines companies (EDBs) compare on cost, reliability, losses and looking after their assets, allowing for how big and how spread out each network is.
Scores at a glance
Each row is a lines company. Click a company row to see what's behind its score.
- Score
- Out of 100, where 100 is best and 50 typical. Size and density are allowed for (asset age, condition and renewal aren't adjusted). The four areas count equally.
- Band
- The range covers the middle 90% of ranks as we randomly vary the weight on each of the four areas. Strong or Weak means that whole range stays in the top or bottom third.
- Groups
- Thirds by ICPs per km of line, which is how the table is listed: sparse below 7.5, intermediate 7.5–12.6, dense 12.6+.
Scores compare each EDB with all EDBs.
Scroll the table sideways for the scores.
| EDB | ICPs/km ICPs per km Connections per kilometre of line. Each home or business connected is an ICP (installation control point).Low means a spread-out, mostly rural network; high means a compact, urban one. It sets the density group. | Band (rank range) Band (rank range) How settled the rank is. We re-rank every EDB 10,000 times, each time weighting the four areas differently; the numbers are the middle 90% of the ranks it gets.Strong or Weak: that whole range is in the top or bottom third. Middle: it overlaps, so treat its position as uncertain. | Overall Overall score The average of the four area scores, out of 100.50 is typical and higher is better, compared with all EDBs. Size and density are allowed for (asset age, condition and renewal aren't adjusted). | Cost25% | Reliability25% | Utilisation and losses25% | Asset management25% |
|---|---|---|---|---|---|---|---|
| Sparse below 7.5 ICPs/km | |||||||
| OtagoNet OtagoNet Poor-condition grades likely on the wrong scale: Pole Mounted Transformer 2024 (0.32% vs usual 39%); Wood poles 2025 (0.0042% vs usual 24%). Without those years: asset score 32 (not 34), rank unchanged. | 4.2 | Strong 1–5 | 74 | 80 | 84 | 96 | 34 |
| The Power Company The Power Company Poor-condition grades likely on the wrong scale: Wood poles 2025 (0.011% vs usual 24%). Without those years: asset score about the same, rank unchanged. | 4.2 | Strong 1–4 | 70 | 80 | 75 | 70 | 54 |
| Scanpower Scanpower No zone substations: takes supply from Transpower at distribution voltage, so step-down losses and costs sit upstream. Flatters its loss ratio and cost.2025 loss ratio 0.4% vs usual 7.3%: implausible, likely a disclosure error. Scored on the five-year median, so no effect. | 6.4 | Strong 4–8 | 63 | 71 | 82 | 43 | 56 |
| Centralines | 4.9 | Middle 6–15 | 57 | 88 | 41 | 41 | 57 |
| Network Waitaki | 7.0 | Middle 7–20 | 52 | 27 | 84 | 48 | 49 |
| The Lines Company | 5.4 | Middle 9–19 | 50 | 38 | 41 | 57 | 64 |
| Firstlight Network | 6.5 | Middle 8–25 | 49 | 88 | 7 | 45 | 56 |
| EA Networks | 6.6 | Middle 10–27 | 43 | 12 | 91 | 18 | 51 |
| Buller Electricity | 7.4 | Middle 19–27 | 39 | 21 | 16 | 59 | 58 |
| Westpower | 6.4 | Weak 25–29 | 29 | 16 | 57 | 16 | 26 |
| Intermediate 7.5–12.6 ICPs/km | |||||||
| Powerco | 12.3 | Strong 1–6 | 66 | 82 | 68 | 50 | 64 |
| Network Tasman | 11.3 | Middle 4–15 | 60 | 98 | 62 | 46 | 32 |
| Northpower | 10.2 | Middle 5–21 | 53 | 57 | 11 | 64 | 79 |
| Waipa Networks Waipa Networks Poor-condition grades likely on the wrong scale: Distribution OH Open Wire Conductor 2024 (0.027% vs usual 4%); Distribution OH Open Wire Conductor 2025 (0.027% vs usual 4%). Without those years: asset score 45 (not 49), rank 17 (not 16).No zone substations (has 33 km of subtransmission): takes supply from Transpower at distribution voltage, so step-down losses and costs sit upstream. Flatters its loss ratio and cost. | 12.4 | Middle 9–23 | 50 | 75 | 12 | 62 | 49 |
| Horizon Networks | 9.7 | Middle 12–26 | 45 | 77 | 27 | 29 | 48 |
| Alpine Energy | 7.7 | Middle 10–26 | 45 | 11 | 84 | 48 | 36 |
| MainPower NZ | 8.4 | Middle 18–26 | 42 | 54 | 43 | 32 | 38 |
| Marlborough Lines | 7.7 | Middle 16–27 | 39 | 7 | 70 | 48 | 30 |
| Top Energy | 8.0 | Weak 27–29 | 26 | 16 | 0 | 43 | 46 |
| Dense 12.6+ ICPs/km | |||||||
| Nelson Electricity Nelson Electricity 91% underground. Reliability isn't adjusted for undergrounding, so the score keeps any benefit; the capital cost shows in totex through a larger asset base (RAB) per unit of network size: $30,653, against a median of $25,103. | 31.3 | Strong 2–5 | 69 | 77 | 86 | 70 | 45 |
| Orion NZ | 18.4 | Middle 3–14 | 60 | 46 | 100 | 43 | 53 |
| Wellington Electricity | 35.9 | Middle 7–17 | 56 | 73 | 79 | 38 | 37 |
| Unison Networks | 12.6 | Middle 7–18 | 51 | 54 | 29 | 73 | 50 |
| WEL Networks | 17.4 | Middle 10–17 | 50 | 45 | 39 | 62 | 55 |
| Electra | 19.5 | Middle 12–17 | 50 | 50 | 43 | 59 | 49 |
| Electricity Invercargill Electricity Invercargill 92% underground. Reliability isn't adjusted for undergrounding, so the score keeps any benefit; the capital cost shows in totex through a larger asset base (RAB) per unit of network size: $32,064, against a median of $25,103. | 26.4 | Middle 16–22 | 46 | 54 | 30 | 59 | 43 |
| Aurora Energy | 15.0 | Middle 15–27 | 40 | 5 | 39 | 43 | 74 |
| Vector Lines | 31.1 | Weak 22–25 | 39 | 23 | 36 | 43 | 54 |
| Counties Energy | 13.2 | Weak 21–28 | 37 | 25 | 14 | 45 | 63 |
· 50 is typical
What stands out
- The top is fairly settled. OtagoNet, The Power Company, Nelson Electricity, Powerco and Scanpower are in the top quarter almost whatever the weighting (97–100% of weightings), and no one beats any of them on all four areas.
- The bottom is pretty settled. Westpower and Top Energy stay in the bottom five however you weight the scores (at least 95% of weightings), and 9–15 EDBs do better than each of them on all four areas.
- Most are too close to call. Only 5 EDBs stay in the top third and 4 in the bottom third across the middle 90% of weightings. The other 20 overlap too much to put in order (a typical rank range is 9 places), so treat them as a group.
- Most all-round. The Power Company has no weak spot: its lowest score is 54 (asset management). Every other top-ten EDB has something at 50 or below.
One EDB against the rest
OtagoNet
OtagoNet A sparse network: 4.2 ICPs per km, 19,312 ICPs, 4,653 km of line, 6% underground.Consumer trust (via The Power Company); price-quality regulated.In the top quarter in 100% of weightings; no EDB beats it on all four areas.Poor-condition grades likely on the wrong scale: Pole Mounted Transformer 2024 (0.32% vs usual 39%); Wood poles 2025 (0.0042% vs usual 24%). Without those years: asset score 32 (not 34), rank unchanged.Showing the top-ranked EDB.
- Overall
- 74 rank 1 of 29
- Band (rank range)
- Strong 1–5
- Within sparse networks
- 67 rank 1 of 10
OtagoNet other EDBs median top quarter of all EDBs typical for its size and density (1.00×) better →
Cost80
Opex per ICP Opex per ICP Running costs (operating spending) per connection a year, in today’s dollars.
$555 as disclosed · log scale
Totex with capital services per √(ICPs × km) Totex with capital services per √(ICPs × km) Running costs plus a yearly charge for the money tied up in the network, per unit of network size, √(ICPs × km). It follows the Commerce Commission’s productivity study by CEPA.
$3,922 as disclosed · log scale
Reliability84
Unplanned SAIDI Unplanned SAIDI SAIDI: the average minutes each customer is without power from unplanned outages a year. Storms are included.
204 min as disclosed · log scale
Unplanned SAIFI Unplanned SAIFI SAIFI: the average number of unplanned outages each customer has a year. Storms are included.
2.11 as disclosed · log scale
Utilisation and losses96
Loss ratio Loss ratio The share of electricity entering the network that is lost on the way, mostly as heat in lines and transformers. The median of the yearly figures.
3.9% as disclosed · log scale
Maximum demand ÷ distribution transformer capacity Maximum demand ÷ distribution transformer capacity Peak demand as a share of the capacity of the transformers that supply homes and businesses. Higher means capacity is better used, up to 0.4; above that counts the same.
0.28 as disclosed · log scale · capped at 0.4
Asset management34
Assets installed before 1980 (45+ years old)
2025 disclosure
Assets in poor condition (grades H1–H2)
Renewal capex ÷ network depreciation Renewal capex ÷ network depreciation Spending on replacing old assets, compared with how fast the network wears out on paper (depreciation). Around 1 keeps pace; above 1.5 counts the same.
capped at 1.5
Totex forecast error Totex forecast error How far actual spending (capital and operating) missed the EDB’s own forecast at the start of the year, over or under, averaged over the years. Lower means steadier planning.
17.9% as disclosed · log scale
Values for every EDB (adjusted)
| EDB | Opex /ICP | Totex /unit | Unplanned SAIDI | Unplanned SAIFI | Losses | Demand ÷ capacity | 45+ years old | Poor condition | Renewal ÷ depn | Forecast error |
|---|---|---|---|---|---|---|---|---|---|---|
| OtagoNet | 0.67× | 0.96× | 0.56× | 0.67× | 0.81× | 1.17× | 23.5% | 11.5% | 0.95 | 1.68× |
| The Power Company | 0.71× | 0.94× | 0.53× | 0.87× | 1.00× | 1.41× | 32.6% | 3.8% | 0.82 | 0.55× |
| Nelson Electricity | 0.77× | 0.96× | 0.56× | 0.56× | 0.94× | 1.06× | 39.6% | 0.3% | 0.56 | 0.85× |
| Powerco | 0.83× | 0.89× | 0.76× | 0.74× | 1.00× | 1.02× | 25.7% | 5.9% | 1.22 | 0.68× |
| Scanpower | 0.91× | 0.91× | 0.73× | 0.42× | 0.96× | 0.94× | 9.7% | 9.9% | 1.04 | 1.03× |
| Orion NZ | 1.04× | 1.00× | 0.29× | 0.37× | 0.93× | 0.89× | 28.7% | 1.3% | 0.72 | 0.89× |
| Network Tasman | 0.70× | 0.66× | 0.77× | 0.81× | 1.09× | 1.06× | 37.1% | 1.7% | 0.85 | 1.99× |
| Centralines | 0.83× | 0.81× | 0.99× | 1.22× | 0.95× | 0.90× | 30.5% | 5.3% | 1.50 | 0.98× |
| Wellington Electricity | 0.87× | 0.92× | 0.61× | 0.72× | 1.12× | 1.05× | 43.2% | 5.0% | 0.96 | 1.06× |
| Northpower | 1.13× | 0.85× | 1.99× | 2.35× | 0.96× | 1.06× | 23.8% | 4.3% | 1.50 | 0.60× |
| Network Waitaki | 1.21× | 1.10× | 0.46× | 0.73× | 0.97× | 0.99× | 24.8% | 5.5% | 1.26 | 1.58× |
| Unison Networks | 1.00× | 0.97× | 2.30× | 0.98× | 0.84× | 1.00× | 26.0% | 13.3% | 1.34 | 0.91× |
| WEL Networks | 1.03× | 1.06× | 1.20× | 1.14× | 0.92× | 1.00× | 26.2% | 0.4% | 0.98 | 1.25× |
| Electra | 1.04× | 0.97× | 0.87× | 1.29× | 1.06× | 1.11× | 30.6% | 3.5% | 1.09 | 1.17× |
| The Lines Company | 1.06× | 1.08× | 1.04× | 1.18× | 1.11× | 1.22× | 28.2% | 3.0% | 1.19 | 0.87× |
| Waipa Networks | 0.95× | 0.83× | 2.02× | 1.46× | 0.91× | 0.99× | 11.7% | 4.2% | 0.85 | 1.70× |
| Firstlight Network | 0.83× | 0.82× | 2.63× | 2.02× | 1.13× | 1.07× | 32.3% | 7.3% | 1.50 | 0.75× |
| Electricity Invercargill | 0.93× | 1.06× | 1.15× | 1.38× | 1.06× | 1.14× | 31.5% | 14.5% | 1.03 | 0.62× |
| Horizon Networks | 0.89× | 0.86× | 1.41× | 1.35× | 1.02× | 0.91× | 28.2% | 7.6% | 0.88 | 0.67× |
| Alpine Energy | 1.39× | 1.13× | 0.72× | 0.50× | 0.85× | 0.80× | 32.7% | 18.9% | 1.50 | 1.08× |
| EA Networks | 1.18× | 1.48× | 0.42× | 0.62× | 1.49× | 0.96× | 5.8% | 4.1% | 0.70 | 1.26× |
| MainPower NZ | 0.97× | 0.98× | 1.40× | 0.91× | 0.95× | 0.80× | 20.1% | 11.1% | 0.72 | 1.16× |
| Aurora Energy | 1.37× | 1.17× | 0.93× | 1.34× | 1.10× | 1.05× | 23.8% | 6.4% | 1.50 | 0.61× |
| Marlborough Lines | 1.30× | 1.22× | 0.73× | 0.78× | 0.82× | 0.79× | 26.4% | 10.8% | 0.63 | 1.03× |
| Vector Lines | 1.06× | 1.17× | 1.49× | 1.13× | 0.98× | 0.97× | 23.2% | 4.8% | 1.12 | 1.57× |
| Buller Electricity | 1.24× | 1.11× | 2.00× | 1.40× | 0.86× | 0.95× | 26.9% | 2.5% | 0.76 | 0.72× |
| Counties Energy | 1.12× | 1.15× | 1.59× | 2.21× | 0.92× | 0.86× | 9.4% | 7.6% | 1.50 | 1.24× |
| Westpower | 1.33× | 1.13× | 0.87× | 0.90× | 1.19× | 0.92× | 11.8% | 24.7% | 0.40 | 1.40× |
| Top Energy | 1.22× | 1.17× | 3.59× | 2.42× | 1.33× | 1.09× | 22.7% | 8.1% | 0.70 | 0.90× |
What good looks like
Figures: median / top quarter, as disclosed, 2021–2025
| Measure | All EDBs29 EDBs | Sparse networksbelow 7.5 ICPs/km, 10 EDBs |
|---|---|---|
| Cost | ||
| Opex per ICP Opex per ICP Running costs (operating spending) per connection a year, in today’s dollars.lower is better | $553 / $399 | $726 / $587 |
| Totex with capital services per √(ICPs × km) Totex with capital services per √(ICPs × km) Running costs plus a yearly charge for the money tied up in the network, per unit of network size, √(ICPs × km). It follows the Commerce Commission’s productivity study by CEPA.lower is better | $4,353 / $3,815 | $4,197 / $3,826 |
| Reliability | ||
| Unplanned SAIDI Unplanned SAIDI SAIDI: the average minutes each customer is without power from unplanned outages a year. Storms are included.lower is better | 170 min / 98 min | 199 min / 113 min |
| Unplanned SAIFI Unplanned SAIFI SAIFI: the average number of unplanned outages each customer has a year. Storms are included.lower is better | 1.62 / 1.13 | 1.86 / 1.29 |
| Utilisation and losses | ||
| Loss ratio Loss ratio The share of electricity entering the network that is lost on the way, mostly as heat in lines and transformers. The median of the yearly figures.lower is better | 5.0% / 4.5% | 6.9% / 5.1% |
| Maximum demand ÷ distribution transformer capacity Maximum demand ÷ distribution transformer capacity Peak demand as a share of the capacity of the transformers that supply homes and businesses. Higher means capacity is better used, up to 0.4; above that counts the same.higher is better | 0.27 / 0.29 | 0.25 / 0.28 |
| Asset management | ||
| Assets installed before 1980 (45+ years old) lower is better | 26.2% / 23.2% | 25.8% / 14.7% |
| Assets in poor condition (grades H1–H2) lower is better | 5.5% / 3.8% | 5.4% / 3.9% |
| Renewal capex ÷ network depreciation Renewal capex ÷ network depreciation Spending on replacing old assets, compared with how fast the network wears out on paper (depreciation). Around 1 keeps pace; above 1.5 counts the same.higher is better | 0.98 / 1.26 | 0.99 / 1.24 |
| Totex forecast error Totex forecast error How far actual spending (capital and operating) missed the EDB’s own forecast at the start of the year, over or under, averaged over the years. Lower means steadier planning.lower is better | 9.0% / 7.0% | 11.8% / 8.8% |
Method
- A comparison, not an efficiency benchmark. It places each EDB among the others on public disclosures, allowing for network size and density. Anything the adjustment misses, like terrain, weather or history, still shows in the scores, so a low score isn’t proof a network is inefficient.
- What’s compared. 10 measures in four areas: cost, reliability, utilisation and losses, and asset management. Each is the EDB’s figure over 2021–2025, in real dollars of the year to June 2025 (CPI). Asset age is a 2025 snapshot, and the loss ratio is the median year, so one odd year doesn’t move it.
- One adjustment for size and density. Cost, reliability, utilisation and losses, and forecast error are each compared with what’s typical for a network of the same size and density. Across all EDBs, each measure is fitted on log ICPs and log km (the same as fitting on network size, √(ICPs × km), and density), and an EDB is scored on its figure divided by that fitted value. 1.00× is typical; 0.80× on a measure where lower is better is 20% better than typical. Losses and transformer use also allow for energy delivered per ICP. Asset age, condition and renewal aren’t adjusted.
- Density groups. Thirds by connections per km of line (sparse, intermediate and dense), so each EDB can also be ranked against networks like it.
- Scores. Each measure becomes a percentile out of 100, where higher is better. An area’s score is the average of its measures, and the overall score is the average of the four areas, so each counts for 25%. Two measures are capped: transformer use above 0.4 and renewal above 1.5× depreciation score the same as the cap.
- Ranges and bands. To test how much the equal weights matter, the ranking is rerun 10,000 times with random weights on the four areas, scattered around equal (each roughly 10–45%). The range is the middle 90% of an EDB’s ranks; Strong or Weak means the whole range is in the top or bottom third.
- Limits. Reliability includes storms and isn’t adjusted for how much of a network is underground. Asset condition is each EDB’s own grading.
- Data. Commerce Commission, electricity distribution information disclosures (CC BY 4.0). Figures are derived and adjusted by NZ Infrastructure Insights and are not endorsed by the Commission. Everything is in the ZIP, with a readme of the sources and columns.