A Working Example
This is a working residential site demonstrating how solar, batteries and simple control can reduce electricity costs, support the network, and enable renters to participate. It introduces the concept of “capacity value” at the LV level.
fig.1 Cycle the battery 1x per day, discharging into peak 07:00 to 11:00, recharging from solar in off-peak (after 11:00). Note: hot water is also scheduled to charge after 11:00 from solar.
The system is installed at a typical home and consists of a 6 kW solar array and an 8 kWh battery. In March it produced 840 kWh of energy while the household consumed 360 kWh. For the trial the battery will be cycled 2x per day. Grid electricity is only used during a controlled off-peak window between 2am and 4am. At all other times the home operates on solar and battery.
fig.2 Cycle battery x2 by charging at low cost time 02:00 to 04:00 and again from afternoon solar. On a good day injecting 4 kW at peak times (7 kWh/day). Over a month this averages 2.5 kW.
The system follows a simple pattern. It charges when electricity is cheap and discharges when electricity is valuable. Overnight the battery charges at low cost. In the morning and evening, when demand on the grid is highest, it discharges to supply the home or export. During the day, solar meets household demand first and excess is exported.
This behaviour can be measured and priced.
fig.3 Typical very low demand and therefore low cost energy around 04:00 daily – likely due to geothermal not ramping down from 100% generation when demand is low.
The Value Stack – Energy + Distribution Costs
This trial introduces a four-layer value stack: Solar generation provides low-cost energy. The battery shifts energy in time. Retail pricing (via Supa) creates arbitrage opportunities. Distribution signals (via PowerCo) reward behaviour that reduces network stress.
The current PowerCo distribution credit (–7c/kWh) is static and does not reflect real-time network conditions. The trial proposes replacing this with simple time-bound credits aligned to peak demand periods.
Local voltage data shows a clear correlation with network stress, with elevated voltage during solar export periods and reduced voltage during peak demand. The trial uses voltage as a proxy for constraint, targeting a flatter profile around 238 V through coordinated charging and discharging.
fig.4 Discharge battery at times when grid requires support and charge to remove over-voltage peaks.
The proposed change replaces the current WACC-based distribution pricing model, which relies heavily on fixed daily charges, with a cost-reflective structure that aligns revenue recovery with actual network usage and constraints. Instead of recovering costs through static charges unrelated to system conditions, distribution pricing is shifted to time and condition-based signals, where higher charges apply during peak periods of network stress and lower charges apply when capacity is available. Critically, consumers are also credited for exporting energy during constrained periods, reflecting the real value of distributed energy resources in reducing peak demand and stabilising voltage. This approach maintains revenue neutrality for the network while incentivising behaviours that “sweat” existing assets, defer capital investment, and enable consumers to actively participate in supporting grid performance through their own solar and battery systems.
Monthly Bill (March) – Cost-Reflective Distribution Model
Energy Only (Retail – Supa)
| Item | kWh | Rate (NZD/kWh) | Amount (NZD) | What it means |
|---|---|---|---|---|
| Off-peak charging (02:00–04:00) | 240 | 0.10 | 24.00 | Cheap energy used to fill battery |
| Morning peak export | 216 | -0.23 | -49.68 | Energy sold back at high price |
| Midday solar export | 264 | -0.17 | -44.88 | Extra solar exported |
| Evening peak battery discharge value | 198 | -0.23 | -45.54 | Battery used during peak pricing |
| Reserve (hold for overnight use ) | 42 | 0.17 | 7.14 | Value of holding energy for later |
| Net Energy | -108.96 | Net credit from energy use |
What is Capacity Value?
Electricity bills are currently based on energy (kWh), but network costs are driven by peak demand (kW). At transmission level this is well understood and managed. At the distribution level it has not been explicitly priced.
This site consistently delivers around 2–3 kW during morning and evening peak periods. This reduces loading on the local transformer and feeder, deferring network upgrades. At approximately $300 per kW per year, this will equate to around $750 per year of value to the site.
This is not a new cost. It is the avoided cost of network investment, now made visible at the edge.
Add Distribution (PowerCo – Cost-Reflective Trial)
If distribution pricing reflects time and capacity rather than fixed charges, the same site delivers additional value.
| Item | Units | Rate | Amount (NZD) | Description |
|---|---|---|---|---|
| Off-peak delivery (02:00–04:00) | 240 kWh | 0.05 | 12.00 | Low-cost network use |
| Peak delivery (import) | 0 kWh | 0.15 | 0.00 | No peak import |
| Peak injection credit | 216 kWh | -0.10 | -21.60 | Export during constraint |
| Capacity payment (peak support) | 2.5 kW | -25.00 | -62.50 | Verified peak support |
| Fixed charge | – | – | 0.00 | Removed |
| Net Distribution Total | -72.10 | Net credit |
“Capacity payment” is a credit for helping the electricity network during peak times. The battery reduces demand when the network is under stress, which avoids expensive upgrades. The network pays for that support based on how much power (monthly average 2.5 kW) the system can reliably provide during peak periods.
Monthly bill comparison
| Item | NZD/month |
|---|---|
| Baseline bill before system | 125.00 |
| Energy credit (Supa) | -108.96 |
| Network credit (PowerCo trial) | -72.10 |
| System rental | 180.00 |
| Net monthly bill after system | -1.06 |
| Monthly improvement vs old bill | 126.06 |
fig.5 System Rental is a charge applied to the renter, paid to the “investor” for use of the system. About 3.5% annual interest. Trial to demonstrate how bill cost ensures renter has incentive to participate. The result is that the system fully offsets the original electricity cost, even when financed.
Changes to EDB Revenue
Under the current WACC-based model, PowerCo recovers a material part of its revenue through fixed daily charges that do not reflect whether a consumer is helping or burdening the network at any particular time. Under a cost-reflective trial model, that fixed charge would be removed and replaced with time-based delivery charges and injection credits that follow actual network conditions. Consumers would pay for using the network when they import during constrained periods, pay much less when importing off-peak, and receive credits when exporting or discharging in ways that reduce peak demand and support local voltage. In this way the revenue stream to the EDB shifts from passive asset recovery toward active network service recovery, rewarding consumers who use their solar, batteries and other controllable assets to flatten voltage, shift load and defer network upgrades.
Annual Return
The site delivers an estimated annual energy value of approximately $1,173 through low-cost overnight charging, morning peak export, evening peak discharge and controlled reserve management. If PowerCo replaces fixed daily charges with cost-reflective off-peak delivery pricing, peak injection credits and a capacity payment, the site would deliver a further $115 per year of distribution value, taking total stacked system value to about $1,288 per year. The home’s pre-solar/battery install electricity bill was approximately $1,500 per year, so the total annual improvement versus the old bill is about $2,788. On an install cost of $22,055, this equates to a simple ROI of 12.6% and a simple payback of 7.9 years.
The trial is initially manual, using scheduled control within the inverter. The next phase is automation, where retailer or network signals directly control device behaviour. This creates a pathway toward a scalable model where residential consumers participate in energy markets as active providers of flexibility.
The outcome will inform future pricing design, including the potential to replace fixed network charges with cost-reflective, time and constraint based pricing aligned with actual system conditions.
The End State
Transmission Capacity
Transpower already applies the concept of capacity value. At that level, the system is planned and operated around peak demand. Generation is dispatched, constraints are managed, and investment decisions are driven by how much power (kW or MW) must be delivered at critical times. The value of capacity is understood because it determines whether the system can meet demand without failure.
This works because the transmission network deals with a small number of large assets. Capacity is visible, measurable, and contractable. The system is designed to ensure that sufficient power is available during peak periods, and the cost of providing that capacity is embedded in pricing and investment decisions.
Extend the Transpower Model to the Edge (Distribution Level)
At the distribution edge, the same physical reality exists, but to date it has not been priced in the same way. Network costs are still largely recovered through daily fixed charges and energy-based tariffs (kWh), even though the drivers of cost are the same as at transmission level: peak demand and the assets required to support it.
The Third Avenue trial proposes to show how this gap can now be addressed. With bi-directional energy flows, smart meters, and controllable devices, households can supply capacity as well as consume energy. This allows pricing to move from a simple kWh model to one that reflects time, location, and capacity.
A standard residential system, consisting of solar and a battery, can deliver approximately 2–3 kW during peak periods. This reduces demand on the local transformer and feeder, directly addressing the same constraint that drives investment at transmission level. Measured over a year, this equates to approximately $750 of value when expressed as capacity at around $300 per kW-year.
This is not a new value. It is the same value that transmission systems already recognise. The difference is that it has not been measured or settled at the distribution level.
Households can now Supply “Capacity Value”
The reason this is now possible is the combination of bi-directional energy flows, smart metering, and controllable devices. Households are no longer passive loads. They can inject energy, shift demand, and respond to signals. This turns each home into a small, controllable node.
Once this is recognised, the pricing model can be extended. Instead of valuing only energy, the system begins to value capacity at the edge. This means defining peak windows, measuring the average power delivered or avoided during those periods, and applying a consistent value per kW. The result is a simple, auditable mechanism that aligns pricing with physical reality.
A More Advanced System
As more devices become controllable, and as pricing reflects time and location, the network begins to behave less like a static infrastructure and more like a coordinated system. Bi-directional energy flows can be shaped rather than simply accommodated. Demand can respond to conditions. Local resources can be prioritised.
This is the foundation of what has been described as an “internet of energy”, a system of many distributed nodes interacting dynamically, rather than a one-way flow from central generation to passive consumption.
This advanced stage introduces greater flexibility in how energy flows physically through the network. Power electronics, including inverters and emerging routing technologies, allow energy to be directed, isolated, or reconfigured in response to events. In the case of a local fault, energy can be rerouted or demand can be reduced through pricing signals, avoiding widespread disruption.
Over time, this leads to the emergence of localised networks that can operate semi-independently, improving resilience. These networks rely increasingly on direct current (DC) within local domains, reducing conversion losses and simplifying control. Many modern devices, including batteries, solar panels, LEDs, TVs and electronics, already operate natively in DC. Aligning the network with this reality improves efficiency.
The Benefits are Cumulative
First, efficiency improves because energy is used closer to where it is produced, reducing losses and the need for long-distance transport. Second, resilience improves because the system can respond dynamically to faults, either through rerouting or controlled reduction in demand. Third, investment efficiency improves because existing assets are used more effectively, delaying or avoiding the need for expansion.
None of this requires a sudden transformation. It can be introduced incrementally.
The Third Avenue site demonstrates that the core elements already exist: measurable capacity at the edge, controllable demand, and price-responsive behaviour. By extending the principles already used at transmission level to distribution, the system can evolve toward a more efficient and resilient model.
This is not a change in physics. It is a change in how we recognise and value what the system is already doing.
Renewable3D and beyond
Energy can now be valued across three dimensions: how much (kWh), when it is delivered, and where it is delivered. Batteries convert energy into capacity by delivering power at the right time and place.
A fourth dimension is emerging. Electricity follows physical paths (routes) defined by network impedance. Energy delivered locally uses minimal infrastructure, while energy travelling further uses more network assets. Pricing that reflects this improves efficiency and fairness.
Internet of energy
As more devices become controllable, the network evolves into a coordinated system. Demand can respond to price signals. Energy can be shifted, prioritised, and, where technology allows, rerouted.
In the event of a constraint or outage, the system can respond in two ways. Pricing signals can reduce demand by encouraging flexible loads to defer consumption. Power electronics and network reconfiguration can redirect flows around the constraint.
Together, these form the basis of an “internet of energy”, where many distributed nodes interact dynamically to maintain balance and resilience.
The result is a network that moves from static capacity built for worst-case scenarios (built to cover peaks) to a dynamic system where capacity is delivered where and when it is needed, forming the basis of a more responsive and ultimately more efficient electricity system.
Conclusion
This trial shows that the technology already works. Households can reduce costs, support the network, and provide measurable capacity. The remaining step is to align pricing with the physical reality of the system.
When that occurs, participation can extend beyond homeowners to include renters and low-income households, ensuring that the benefits of the energy transition are shared broadly.
References
Joint EA, ComCom and EECA letter – call to EDBs for efficient Non-Network Solutions https://www.ea.govt.nz/documents/9322/Joint_letter_to_distributors_-_non-network_solutions.pdf
Jonas Birgersson “Internet of Energy” discussed here https://solarenergy.kiwi/energy-internetification/
EnergyNet Full version https://arxiv.org/pdf/2509.08152
PowerCo, 7c/kWh in winter peaks, a small first step! pricing schedule