Case study: Water-energy nexus optimization — a city or utility pilot and the results so far

The city of Windhoek, Namibia, spends roughly 45% of its municipal electricity budget on water treatment and distribution, a ratio that climbs to 60% during drought years when the city activates its direct potable reuse facility and emergency borehole network. In 2023, the City of Windhoek partnered with the Deutsche Gesellschaft fuer Internationale Zusammenarbeit (GIZ) and Namibia Water Corporation (NamWater) to launch a water-energy nexus optimization pilot that targeted a 25% reduction in energy consumption per cubic meter of water delivered. After 18 months of operation, the pilot has achieved a 19% reduction in specific energy consumption across the treatment and distribution chain, saving approximately $2.1 million annually in electricity costs while maintaining water quality compliance across all 14 distribution zones.

Why It Matters

Water and energy systems are deeply interdependent. The International Energy Agency estimates that water-related energy consumption accounts for 4% of global electricity demand, a figure projected to rise to 6% by 2040 as utilities turn to more energy-intensive sources such as desalination, deep aquifer pumping, and advanced treatment for potable reuse (IEA, 2024). In emerging markets, this interdependence is particularly acute. Sub-Saharan African utilities lose an average of 37% of treated water to leaks and non-revenue water, meaning that more than a third of the energy embedded in water treatment is effectively wasted (World Bank, 2025).

Windhoek is a critical test case because it faces conditions that are becoming common across arid and semi-arid regions worldwide: declining surface water availability, growing population pressure, aging infrastructure, and electricity costs that have risen 140% in real terms over the past decade. The city already operates the world's longest-running direct potable reuse facility, the New Goreangab Water Reclamation Plant, which has been treating wastewater to drinking water standards since 1968. What the city lacked was an integrated optimization framework that treated water and energy as a single system rather than separate operational silos.

Key Concepts

The water-energy nexus refers to the bidirectional relationship between water and energy: energy is required to extract, treat, and distribute water, while water is needed to generate energy through cooling, hydropower, and fuel extraction. Optimization at this nexus involves reducing energy consumption in water systems, reducing water consumption in energy systems, or both simultaneously.

Specific energy consumption (SEC) measures the electricity required to deliver one cubic meter of treated water to the customer, typically expressed in kWh per cubic meter. Windhoek's pre-pilot SEC averaged 2.8 kWh per cubic meter across its blended supply portfolio, which includes surface water from the Von Bach and Omatako dams (0.6 kWh/m3), reclaimed water from the Goreangab plant (1.4 kWh/m3), and borehole extraction from the Windhoek Managed Aquifer Recharge scheme (1.8 kWh/m3). The blended SEC rises sharply during droughts when the city shifts toward higher-energy sources.

Non-revenue water (NRW) represents the volume of treated water that enters the distribution system but generates no income, due to leaks, metering errors, or theft. Each cubic meter of NRW carries the full energy cost of treatment and pumping but produces zero economic return, making NRW reduction one of the highest-impact interventions for water-energy optimization.

What's Working

Integrated SCADA and Energy Management

The pilot's centerpiece is the integration of Windhoek's existing supervisory control and data acquisition (SCADA) system with a new energy management layer developed by South African firm WEC Projects. The system collects real-time data from 287 flow meters, 94 pressure sensors, 38 pump stations, and 12 reservoir level monitors, correlating water flow data with electricity consumption from 156 power meters installed across the treatment and distribution network.

The integrated platform enables time-of-use pumping optimization, shifting energy-intensive operations such as reservoir filling and high-lift pumping to off-peak electricity tariff periods. NamPower, Namibia's electricity utility, charges industrial customers 40% less during off-peak hours (22:00 to 06:00) compared to peak periods (07:00 to 10:00 and 18:00 to 20:00). By reprogramming pump schedules to maximize off-peak operation while maintaining minimum reservoir levels, the pilot reduced peak-period electricity purchases by 34% in the first year, translating to $870,000 in annual tariff savings alone.

Variable Speed Drive Retrofits

The pilot replaced 22 fixed-speed pumps with variable speed drive (VSD) units across five critical pump stations, prioritized based on energy audit data showing which stations had the highest gap between actual and theoretical minimum energy consumption. Pump stations serving the Windhoek Northern Industrial Area and Katutura residential zones showed the greatest inefficiency, with actual energy consumption 45 to 65% above the theoretical minimum due to throttle-valve flow control on oversized pumps.

After VSD installation, these stations reduced specific energy consumption by 28 to 42%, with an average payback period of 2.3 years. The total capital investment for the VSD retrofit program was $1.4 million, financed through a GIZ concessional loan at 2% interest with a 7-year repayment term.

Pressure Management and Leak Reduction

Windhoek's distribution network operates across an elevation range of 240 meters, creating significant pressure differentials between hilltop and valley zones. Pre-pilot, the city maintained uniform high pressure to ensure adequate supply to hilltop customers, resulting in excessive pressure (and correspondingly higher leak rates) in lower-elevation zones.

The pilot installed 18 pressure-reducing valves (PRVs) with real-time telemetry in a phased rollout designed by consulting firm Miya (now part of Veolia). The PRVs modulate downstream pressure based on demand patterns, maintaining minimum service pressure of 2.0 bar at all customer connections while reducing average network pressure by 22%. The pressure management program reduced burst frequency by 31% in the first 12 months and cut NRW from 24.6% to 19.8%, recovering approximately 2,400 cubic meters per day of previously lost treated water. At Windhoek's blended SEC of 2.3 kWh per cubic meter (post-optimization), this NRW reduction saves 5,520 kWh per day in embedded energy.

Solar-Powered Borehole Pumping

Eight of Windhoek's 23 managed aquifer recharge boreholes were equipped with solar photovoltaic arrays and battery storage systems, eliminating grid electricity consumption for borehole extraction at these sites. Each installation comprises a 30 kW solar array, 60 kWh lithium iron phosphate battery bank, and intelligent charge-pump controller that optimizes extraction volumes based on solar availability, aquifer levels, and system demand forecasts.

The solar borehole program was implemented by Namibian firm SunCycle Energy at a total cost of $960,000 for all eight installations. Grid electricity savings at these boreholes total approximately $185,000 per year, yielding a simple payback of 5.2 years. More importantly, the solar boreholes provide drought-resilient water supply independent of grid reliability, which has become a strategic priority following NamPower's load-shedding events in 2024 that interrupted grid-dependent pumping for up to 6 hours per day.

What's Not Working

Data Integration Challenges

Integrating legacy SCADA systems (some dating to the 1990s) with the new energy management platform proved far more complex than anticipated. Windhoek's water infrastructure uses at least four different SCADA protocols (Modbus, DNP3, OPC-DA, and proprietary Allen-Bradley systems), and approximately 30% of existing sensors produced data in formats incompatible with the new platform without custom middleware. The data integration phase consumed 14 months of the project timeline against an original estimate of 6 months, and the middleware development budget exceeded initial estimates by $320,000.

Several critical data gaps remain. Flow meters at 23 of 287 monitoring points produce readings with greater than 10% measurement uncertainty, limiting the accuracy of water balance calculations in those zones. The project team has deferred replacement of these meters to Phase 2 due to budget constraints.

Institutional Coordination Barriers

The pilot exposed significant coordination challenges between the City of Windhoek (responsible for water distribution), NamWater (responsible for bulk water supply), and NamPower (responsible for electricity supply). Optimizing pumping schedules for energy cost reduction requires access to day-ahead electricity pricing signals, which NamPower was initially unwilling to share in real time due to concerns about market-sensitive information. It took 8 months of negotiation to establish a data-sharing agreement that provides 24-hour-ahead pricing forecasts to the optimization platform.

Similarly, coordinating reservoir levels between NamWater's bulk supply reservoirs and the City's distribution reservoirs requires operational trust and compatible control systems that do not yet fully exist. During a 3-week period in August 2025, miscommunication between NamWater's bulk supply scheduling and the City's optimized pump scheduling resulted in a distribution zone running below minimum pressure for 6 hours, affecting approximately 12,000 households.

Limited Applicability of Solar Boreholes in Urban Zones

While solar-powered borehole pumping works well for peri-urban and rural boreholes with moderate extraction rates (50 to 150 cubic meters per day), the approach proved impractical for Windhoek's high-capacity urban boreholes that extract 500 to 1,200 cubic meters per day. The solar array and battery sizes required for these volumes would cost $450,000 to $800,000 per installation with payback periods exceeding 12 years, making them uneconomic compared to grid-powered VSD pumps optimized for off-peak operation. The project team now recommends solar borehole installations only for sites with extraction rates below 200 cubic meters per day.

Key Players

Established Organizations

• City of Windhoek: Municipal water authority operating treatment, reclamation, and distribution for Namibia's capital city with 450,000 residents
• NamWater (Namibia Water Corporation): National bulk water supplier managing dams, pipelines, and inter-basin transfers across Namibia
• Veolia (Miya): Global water services firm providing pressure management consulting and NRW reduction expertise across 40 countries
• GIZ: German development agency providing $4.2 million in technical assistance and concessional financing for the pilot
• Grundfos: Danish pump manufacturer supplying VSD-equipped pump units and remote monitoring for the retrofit program

Startups and Regional Firms

• WEC Projects: South African water and energy engineering firm that developed the integrated SCADA-energy management platform
• SunCycle Energy: Namibian solar installer specializing in off-grid water pumping systems for municipal and agricultural clients
• Utilis (Israel): Satellite-based leak detection provider whose synthetic aperture radar analysis identified 47 previously unknown leaks in Windhoek's trunk main network

Investors and Development Partners

• KfW Development Bank: Provided EUR 8 million in co-financing for Windhoek's broader water infrastructure modernization program
• African Development Bank: Funding a $15 million extension of the managed aquifer recharge scheme that complements the nexus optimization pilot
• European Investment Bank: Committed EUR 12 million to NamWater's bulk supply efficiency program, which includes pumping optimization on the Von Bach pipeline

KPI Summary

KPI	Pre-Pilot	Post-Pilot (18 months)	Target (36 months)
Specific energy consumption (kWh/m3)	2.8	2.27	2.1
Non-revenue water (%)	24.6	19.8	16.0
Peak electricity purchases (% of total)	58	38	30
Burst frequency (events/100 km/yr)	42	29	22
Annual electricity cost ($M)	11.2	9.1	8.2
Solar borehole share of extraction (%)	0	14	22
Average network pressure (bar)	4.8	3.7	3.5

Action Checklist

• Conduct a baseline energy audit across all water treatment and distribution assets, establishing SEC benchmarks for each supply source and pump station
• Map time-of-use electricity tariff structures and quantify potential savings from shifting pump schedules to off-peak periods
• Prioritize VSD retrofits based on the gap between actual and theoretical minimum energy consumption at each pump station
• Implement pressure management with real-time telemetry in distribution zones with elevation differentials exceeding 50 meters
• Establish data-sharing agreements with bulk water and electricity suppliers before beginning integrated optimization
• Deploy solar-powered pumping at low-to-medium capacity extraction points (below 200 cubic meters per day) where grid reliability is poor
• Install NRW monitoring infrastructure including district metered areas and continuous flow balance reporting
• Develop institutional coordination protocols with clear escalation procedures for inter-utility scheduling conflicts

FAQ

Q: What is the typical payback period for water-energy nexus optimization investments in emerging market utilities? A: Payback periods vary significantly by intervention type. Operational optimization (pump scheduling, time-of-use shifting) delivers returns within 6 to 12 months with minimal capital investment. VSD pump retrofits typically pay back in 2 to 4 years depending on electricity tariff structures and pump utilization rates. Pressure management programs show payback periods of 3 to 5 years when accounting for reduced burst repair costs and NRW recovery. Solar borehole installations in suitable locations achieve payback in 4 to 7 years. Windhoek's blended portfolio of interventions is tracking toward a weighted average payback of 3.1 years across all capital investments.

Q: How transferable is the Windhoek model to other emerging market cities? A: The core optimization approach is highly transferable, but several Windhoek-specific conditions should be noted. Windhoek benefits from relatively modern SCADA infrastructure by Sub-Saharan African standards, a well-established institutional framework for water reclamation, and electricity tariffs with significant peak-to-off-peak differentials (40%). Cities without existing SCADA systems will face higher upfront costs for instrumentation and connectivity. The World Bank estimates that establishing baseline monitoring infrastructure for a 500,000-person city in Sub-Saharan Africa costs $3 to $7 million depending on network complexity. Cities in South Asia and Latin America with similar water stress but different tariff structures should model their specific economics before committing to capital programs.

Q: What role does artificial intelligence play in water-energy nexus optimization? A: Windhoek's pilot uses rule-based optimization algorithms rather than AI or machine learning, a deliberate choice driven by the need for operator transparency and regulatory acceptance. However, the data infrastructure being built provides the foundation for future AI deployment. Utilities in higher-income markets such as Singapore's PUB and Israel's Mekorot are already using machine learning for demand forecasting, leak prediction, and real-time pump scheduling optimization, reporting 5 to 15% additional energy savings beyond what rule-based systems achieve. The GIZ project team plans to pilot a machine learning demand forecasting module in Phase 2, beginning in 2027, after establishing sufficient historical data for model training.

Q: How does NRW reduction compare to supply-side efficiency as an energy-saving strategy? A: In systems with NRW above 20%, reducing losses is almost always more cost-effective per kWh saved than improving treatment or pumping efficiency. Each cubic meter of recovered NRW carries the full embedded energy of treatment and distribution (2.3 kWh/m3 in Windhoek's case) but requires only the marginal cost of leak repair or pressure management to recover. Windhoek's pressure management program cost $1.1 million and saves approximately 2,015 MWh per year in embedded energy, equating to $0.55 per kWh of annual savings. By comparison, the VSD retrofit program cost $1.4 million and saves approximately 1,460 MWh per year, equating to $0.96 per kWh of annual savings. Both are strong investments, but NRW reduction delivers faster returns in high-loss systems.

Sources

• International Energy Agency. (2024). Water-Energy Nexus: World Energy Outlook Special Report. Paris: IEA.
• World Bank. (2025). Reducing Non-Revenue Water in Sub-Saharan Africa: Benchmarks and Best Practices. Washington, DC: World Bank Group.
• City of Windhoek. (2025). Water-Energy Nexus Optimization Pilot: 18-Month Progress Report. Windhoek: Department of Infrastructure, Water, and Technical Services.
• GIZ. (2025). Integrated Water-Energy Management in Southern African Cities: Windhoek Pilot Phase 1 Results. Eschborn: Deutsche Gesellschaft fuer Internationale Zusammenarbeit.
• NamWater. (2024). Bulk Water Supply Efficiency Program: Annual Review 2024. Windhoek: Namibia Water Corporation.
• Grundfos. (2025). Variable Speed Drive Pumping in Municipal Water Systems: Energy Savings Case Studies from Sub-Saharan Africa. Bjerringbro: Grundfos Holding.
• SunCycle Energy. (2025). Solar-Powered Borehole Pumping: Performance Data from Windhoek Managed Aquifer Recharge Network. Windhoek: SunCycle Energy (Pty) Ltd.
• Veolia. (2025). Pressure Management and NRW Reduction in Arid-Climate Distribution Networks. Paris: Veolia Environnement.