South Africa’s connected water heaters are becoming a test of whether insurers can prevent household losses while helping customers manage electricity costs. Nedbank’s Senseable Geyser programme offers an example of that approach, with technology provider Sensor Networks reporting approximately R50 million in estimated energy savings over 18 months.
The programme combines monitoring, automatic water shut-off and an app that allows households to manage their electric water heaters, commonly known as geysers in South Africa. According to Sensor Networks, the Nedbank deployment has grown to approximately 10,000 connected geysers and detected and contained close to 2,000 leak events.
Those figures suggest a business case extending beyond the sale of a smart-home device. Customers may benefit from lower energy consumption and less disruption, while insurers have an opportunity to limit damage before a household incident becomes an expensive claim. Understanding that business case requires separating measured consumption from estimated savings, and observed leak interventions from claims that might otherwise have occurred.
How the R50 million estimate was calculated
In written responses to AfricaBusiness.com, Mark Allewell, CEO of Sensor Networks, explained that the system receives energy-consumption updates from its hardware every five minutes. Actual consumption is compared with a reference scenario: a conventional geyser maintained at 65°C without a heating schedule.
The difference in electricity consumption is converted into rand using a tariff of R3.20 per kilowatt-hour. Sensor Networks confirmed that this tariff underpins both the aggregate savings estimate and its reported average monthly household saving of R402.
At that rate, R402 represents approximately 125.6 kWh of estimated electricity savings per household per month. These savings are calculated against the programme’s reference scenario; they should not be interpreted as a verified reduction of the same amount on every customer’s electricity bill.
The comparison matters because households have different starting points. A customer already managing heating schedules carefully may have less scope to reduce consumption than someone whose geyser previously remained at the default setting throughout the day. Actual financial outcomes also depend on the household’s electricity tariff and usage patterns.
Sensor Networks said the active installed base averaged approximately 6,500 to 7,000 geysers during the 18-month period as installations expanded towards 10,000. Its illustrative calculation uses 6,900 devices, multiplied by R402 per month over 18 months, producing approximately R49.93 million.
This explains how the reported total relates to the growing fleet. The underlying consumption dataset and reference model have not been independently audited by AfricaBusiness.com.
What customers pay and what they may gain
According to Allewell, Nedbank subsidises the service for participating policyholders, who pay a monthly subscription of R99. Deducting that fee from the reported R402 average energy saving produces an estimated monthly benefit of R303.
That calculation provides an indication of the customer economics, although individual outcomes remain dependent on actual consumption, electricity tariffs and the company’s savings methodology.
The Sensor Connect app allows customers to monitor consumption, set heating schedules, adjust settings and switch the geyser off remotely. Allewell said the company also uses prompts and guidance within the app to encourage changes in behaviour, such as switching off unnecessary heating during extended absences.
Installing a connected controller creates the ability to manage consumption, but the resulting savings also depend on how the system is configured and used. The service combines hardware, software and customer behaviour.
Containing leaks before damage spreads
The insurance proposition centres on detecting escaping water and closing the supply valve. Sensor Networks says its system can trigger automatic isolation within 10 seconds of detecting a leak.
An alert is also sent to an incident-management platform so that a plumber can be dispatched with information about the problem. This connects detection with a practical response, reducing the need for the policyholder to identify the fault and organise assistance independently.
Automatic shut-off limits the continuing supply of water to a leak, potentially reducing consequential damage while a plumber is dispatched.
Sensor Networks estimates that approximately 40–50% of the roughly 2,000 intercepted events could otherwise have developed into severe claims involving consequential water damage. That corresponds to around 800–1,000 potentially serious incidents.
Asked about the basis for this estimate, Allewell said the company worked with an external data-science firm to analyse four years of historical insurance claims. The analysis used payout amounts as a measure of severity and compared historical claims patterns with data from properties fitted with the technology.
The underlying study was not supplied to AfricaBusiness.com. The 800–1,000 figure is therefore presented as the company’s estimate of potentially severe incidents avoided, rather than an independently verified outcome.
Specific reductions in claims frequency, severity and total claims costs were not disclosed. Verified aggregate water-saving volumes were also not provided, so the programme’s wider insurance and water benefits cannot yet be quantified from the information available.
Testing whether geysers can support demand management
Beyond individual homes, Sensor Networks is testing whether groups of connected geysers can help shift electricity demand away from peak periods.
Allewell clarified that the demand-management programme with Ariston is in a commercial beta phase involving approximately 1,000 connected households. It is not yet a full commercial rollout.
The company cites typical heating-element ratings of 3–4 kW, giving the test fleet a combined connected capacity of approximately 3–4 MW. This describes the aggregate rated power of the equipment, rather than a demonstrated reduction in grid demand.
Actual demand that can be shifted at a particular moment depends on how many heaters would otherwise be operating, household hot-water requirements and the programme’s control arrangements. No measured peak-demand reduction was supplied for this feature.
The trial addresses an important distinction in energy management: reducing total consumption and changing when electricity is consumed are different outcomes. A successful programme needs to demonstrate both the flexibility available and whether it can be delivered while meeting households’ hot-water needs.
The installation network is part of the product
Scaling connected-home technology requires people who can install, maintain and troubleshoot it. Sensor Networks reports that more than 400 plumbers have completed its online training modules, supported by ongoing technical guidance and dedicated support groups.
This points to a practical role for established tradespeople in the deployment of connected devices. The work combines conventional plumbing with controllers, sensors and digital fault information.
The training figure does not establish how many additional jobs have been created or how much participating plumbers’ earnings have increased. It does, however, illustrate why the service network matters: a sensor can report a problem, but somebody still needs to resolve it.
Allewell said the company is working with Ariston towards integrating smart controls at the manufacturing stage. Factory integration could simplify deployment, although the timetable and scale of a broader rollout remain to be demonstrated.
Can the model reach more African households?
Insurance provides a potential route to wider adoption because the organisation funding or subsidising the technology may also benefit from lower claims costs. If avoided losses justify the expenditure, prevention can become part of the insurance service rather than a separate household purchase.
However, that route does not automatically reach households outside formal insurance coverage. Extending the model to lower-income customers would require a workable approach to hardware costs, installation, maintenance and ongoing service fees.
Sensor Networks reports that it already operates in Namibia and is assessing additional regional markets with Ariston. The commercial case in each market would need to reflect local electricity costs, housing conditions, insurance participation and the availability of technical support.
The Nedbank deployment shows how connected devices can bring insurers closer to the events that generate claims. Its next test is whether the reported benefits remain convincing as the fleet grows: through transparent savings calculations, comparable claims outcomes and reliable service for customers.
For African insurers and technology businesses, the opportunity lies in making prevention measurable and repeatable. The strength of the model will depend on whether household savings and avoided damage consistently justify the cost of delivering it.
Reporting note: This article draws on a Sensor Networks press statement and written responses attributed to CEO Mark Allewell, including methodological clarifications supplied to AfricaBusiness.com. Programme figures and projected benefits are company-reported unless otherwise indicated.