Can’t afford a home battery? How a hot water service can be just as good

When I bought my first home many years ago the owner informed me that the electric hot water storage (HWS) system automatically started heating at midnight, coinciding with the night off-peak tariff, providing significant savings on the hot water bill. It seemed like a great idea at the time.

If I had been able to look forward then to now, I would have found that not much has changed. Night-time, off-peak hot water tariffs are still in place, providing coal-fired generators with needed night demand to flatten daily generation.

However, in more recent times there have been big changes in household energy supply. The introduction of rooftop solar has been remarkably successful. But the continued current growth of rooftop solar and unused household solar exports is unsustainable, having already created major stability issues for the grid.

The problem is becoming sufficiently serious that remote methods of controlling energy flows in domestic hot water systems (HWS) to enable their use for storing energy are being investigated.

A recent project entitled ‘Active Hot Water Control’, aims to reduce energy demand for hot water at peak times, to help manage grid stability and lower costs. The project intends to conduct trials on household HWS aggregated  into a VPP that remotely integrates with home energy management systems.

Pending the outcome of this project and others, perhaps the role of household HWS is about to change in the energy plan of the future.

Here, we explore what should be achievable now and what is possible in the future with improved control strategies.

Figure 1 shows the energy flows in a four-person household with rooftop solar but no HWS. Curve (a) is the daily load profile with morning and late afternoon/evening peaks generally considered to be representative of household energy use.

In this example, the household is assumed to use 20kWh per day, including 6kWh (30%) for hot water heating. The hot water usage is assumed to follow the same use profile as the other household loads.

The power from a 4kW PV array on a clear October day in Perth is shown by curve (b). This energy totals 23.3kWh over the full day, significantly greater than the daytime household energy use of 7.5kWh.

Since the solar power exceeds the household load during daytime hours, there is no need for backup power from the grid during the day (curve (c)). However, backup energy totaling 12.5kWh from the grid is still required once the sun has set. Curve (d) shows the excess PV power not used by the household during daytime. Without any storage this energy (15.8kWh) is exported to the grid and attracts a small feed in tariff (FiT) to the householder.

Of the energy flows in Fig1, the backup and excess energies collectively equal the in front of the meter (IFM) energy flow. The remaining flows are all behind the meter (BTM) and not directly measured by the electricity retailer.

The excess power does not contribute to any household power requirement and, with no storage, has become a major source of low-cost daytime power for the grid.

The IFM power flow is, so to speak, the household’s elephant in the room, with its characteristic duck profile associated with the excess household power generated on clear days and rapid ramping of dispatchable power required at sunset to make up for the loss of rooftop solar and increase in household load.

The role of HWS in household energy supply has often been overlooked. However, despite only providing one-way electrical storage, the contribution of HWS to household load shifting and energy storage is important.

Figure 2 shows the household power flows in Fig 1 after the addition of 10kWh storage capacity HWS (~ 1.6 days HWS capacity). The curves (a) – (d) retain their identity defined Fig 1. Curve (e) in Fig 2 shows the PV power used for charging the HWS system.

It is assumed that HWS charging starts in the morning when excess power from the  PV array becomes available. The maximum power for charging the HW is assumed to be 3.5 kW.

On this day, the HWS system is fully charged after ~3 hours. Once fully charged, the power to the HWS is limited to making up for hot water usage during the remainder of the day, using available excess power.

Since the PV power supplied to the HWS would otherwise be excess if there was no HWS, the HWS reduces the  exported energy by6 kWh, not including in/out efficiency losses, taken to be 10%.

The PV power supplied to the HWS is behind the meter (BTM) and increases household energy self-sufficiency in this example from 36.8% to 57%.

Because of the relatively large storage capacity of the HW tank, variations of daily hot water usage pattern on individual days have little effect on system performance.

Despite the large storage capacity, the average daily energy stored and used is limited to the 6kWh HW load. Once the hot water tank is fully charged there is still 11.4kWh of excess energy as compared to 15.8kWh with no HWS. The excess energy can be exported to the grid or stored in batteries.

The use of a heat pump to enhance the performance of a solar HWS is a much-discussed option. Heat pumps typically exhibit COP (Coefficient of  Performance) values of ~ 3. This means that the power delivered to a HWS with a  COP=3 heat pump is three times greater than without a heat pump.

Figure 3 shows the result of adding a 3:1 heat pump to the system in Fig 2. While the heat pump  increases the power from the array by a factor of 3, the daily HW load limit of 6kWh remains the same. The net effect is that a system with a COP=3 heat pump charges 3 times faster but does not deliver any more energy over the day.

Without adding an additional hot water load (or reducing the array size), the additional 4kW power from the heat pump will simply add to the exported power, clearly not the desired result and of no benefit to the householder apart from a small increase in FIT.

Yearly calculations of the effect of array size and HWS on energy delivered are shown in Fig 4. With a 4kW array the HWS increases the energy delivered by 1.5MWh (65%) relative to a system with no HWS.

It can be shown that the addition of the HWS system to a 4kW array is equivalent to adding a 5kWh battery to a  household with no HWS.

The heat pump effectiveness decreases with increasing array size for arrays larger than 2kW. For array sizes greater than ~4 kW the energy supplied to the hot water load, with or without heat pump, is limited by the 6kWh  daily hot water load. As previously noted, any additional power generated by the PV array is excess and exported to the grid.

Yearly values of energy supplied, and tariff cost savings are summarised in Table 1. The savings are relative to a $2202 yearly flat tariff for a 20kWh load with no solar or storage. Adding 4kW rooftop array with no HWS saves the householder $760 a year energy costs. Adding a 10kWh daytime charged HWS to the 4kW array saves a further $432 a year.

For array sizes greater than 4 kW a heat pump HWS provides only very minor additional energy or savings.

To summarize, our calculations show that HWS systems provide very useful energy savings and self-sufficiency for households, provided daytime charging is used. HWS are used by around 50% of Australian households. However, little data on the take-up of daytime solar charging of solar HWS relative to off-peak night charging is available.

From the viewpoint of the householder it is difficult to see any advantage for external, IFM remote control. HWS powered with rooftop solar already operate close to optimum, particularly on clear days, with little or no need for additional controls apart from a daytime charging switch. Nearly all energy is provided by PV.

For example, in Perth, we calculate that 96% of yearly HW load is generated with a 4kW array and 99% with a 6kW array. The energy is behind the meter and, as such, is free of tariffs. Solar PV HWS and daytime charging should be strongly promoted.

However, from the standpoint of grid stability, there is a very definite advantage if HW charging is shifted to the afternoon.

As illustrated in Fig 2, during charging the excess energy available for export is reduced to zero. Once the HWS is charged, the power exported to the grid abruptly increases to nearly the same value it would have without a HWS.

With morning charging this provides no benefit to stability issues associated with rapid ramping of dispatchable generators in the late afternoon.

If HWS charging is shifted to the afternoon, then excess energy exported to the grid will be shifted to the morning, eliminating or significantly reducing any contribution of household export to afternoon grid stability issues.

Comparison of morning versus afternoon HWS charging (noon start time) on yearly values of afternoon export energy is shown in Fig 5. With a 4kW array and 6kWh HW load, afternoon charging reduces afternoon excess energy by a factor of nearly 10 relative to HW charging during the morning.

These results show that HWS can play an important role in stabilising the grid, even with only one-way electrical power flow.

It is noted that the results here for HWS systems are very similar to that in our previous analysis of the effect of household battery charge profile on the distribution of export energy.

While further simulations are needed, particularly on combined HWS and battery systems, afternoon charging, whether it be battery or HWS charging, should be highly effective for late afternoon grid stabilisation.

Paul McCormick is Emeritus Professor Mechanical Engineering at the University of Western Australia