Geothermal Heat in East Africa: Matching Industrial Demand, Resource Risk and Investment

EXECUTIVE THESIS

• Direct-use geothermal heat deserves a distinct investment assessment rather than treatment as a by-product of electricity generation.
• Kenya provides an institutional and technical starting point; Tanzania and Uganda require different resource, demand and regulatory tests.
• The strongest initial proposition is a verified heat service for identifiable customers, with expansion conditional on measured performance.

The central question is whether East African geothermal developers should finance more heat infrastructure now, and what evidence should precede commitment. StoneComms's judgement is conditional: advance specific resource-and-customer combinations, not a general industrial-park promise. A smaller project with measured demand and a credible operating contract may justify investment before a larger resource with no committed buyer.

This is a project-screening argument, not a finding that a particular site is bankable. The public evidence establishes regional policy attention, differentiated country readiness and a dedicated preparation facility. It does not supply the well tests, customer load curves or commercial terms required for an investment decision.

WHY THIS MATTERS NOW

The African Rift Geothermal Conference took place in Moroni from 19 to 24 October 2026. Its programme included direct use, financing, resource development and local livelihoods. GDC's account published on 26 October records participation and technical exchange; neither a conference nor a partnership announcement establishes a financed project. Sources: https://argeo-c11.com/ and https://www.gdc.co.ke/blog/au-revoir-moroni-new-connections-and-geothermal-insights-from-a-successful-argeo-c11/

There is a more substantive implementation signal. AFD's September 2025 agreement with Kenya links Kenyan geothermal experience with French direct-heat expertise and regional cooperation involving Tanzania and Uganda. GDC's July 2026 account describes a technical-assistance programme running from May 2025 to May 2027, including market-readiness work and a January 2026 mission examining a direct-use industrial park at Menengai. These are preparation activities, not evidence of completed industrial delivery. Sources: https://www.afd.fr/en/press-releases/east-africa-poised-renewable-energy-boost-kenya-france-geothermal-agreement and https://www.gdc.co.ke/blog/france-taps-on-gdc-to-unlock-eas-geothermal-potential/

The decision window is therefore practical. Governments can use current preparation work to establish a reproducible appraisal standard before prospective parks and heat networks harden into public commitments. The alternative is to allow a resource estimate, a list of possible applications and an attractive site plan to stand in for a commercial case.

ELECTRICITY AND HEAT ARE DIFFERENT PRODUCTS

IRENA distinguishes direct heating and cooling from electricity generation. Direct uses include greenhouse heating and other established thermal applications, while electricity conversion generally needs a different resource and technology configuration. Technical maturity at global level should not be mistaken for commercial readiness at an individual East African site. Source: https://www.irena.org/Energy-Transition/Technology/Geothermal-energy

Electricity can enter a network that serves many customers. Heat must normally reach a physically compatible user through a local delivery system. That creates a joint siting problem: the resource, the pipe route and the industrial process must work together. Moving a factory to a heat source may increase freight, labour, water or market-access costs enough to erase the thermal advantage.

Nor is every hot fluid suitable for every process. An engineering assessment must establish deliverable temperature, pressure, flow, chemistry and operating stability, together with the requirements of the customer's equipment. A process needing reliable high-temperature steam cannot be appraised using an undifferentiated estimate of underground heat.

Where power and heat share a field, the relevant question is the incremental value of each use under the reservoir-management plan. Heat is not automatically free because a power plant already exists. Extraction, treatment, reinjection, pumping, maintenance and any effect on electricity output must be included in the comparison. These are appraisal requirements proposed by StoneComms, not cost findings for Menengai.

THREE MARKETS, THREE ENTRY CONDITIONS

Kenya: use experience without assuming demand

Kenya's advantage is more than geology. The Menengai electricity model separates public resource development from privately developed generation. AfDB's January 2026 financing announcement describes GDC as the steam supplier for a 35 MW power project, with Kenya Power as the electricity buyer under a 25-year agreement. This illustrates an allocation of responsibilities; it does not prove that an industrial heat customer can use the same contract or credit structure. Source: https://www.afdb.org/en/news-and-events/press-releases/african-development-bank-approves-165-million-loan-boost-kenyas-clean-energy-transition-90373

For direct heat, the customer may be a dairy processor, drying operation or group of smaller firms rather than a national utility. Appraisal must establish operating schedules, seasonal throughput, equipment compatibility and the customer's ability to pay. The relevant counterfactual is the actual alternative at that factory, including its reliability, financing and conversion costs.

The Menengai preparation work offers a useful place to test this discipline. It should publish the distinction between expressions of interest, independently assessed loads and binding commitments. A list of interested tenants is valuable market intelligence, but cannot support debt service on the same terms as a credible long-term purchase obligation.

Tanzania: protect the distinction between a prospect and a service

TGDC's portfolio describes Ngozi as being in resource confirmation, with a stated 70 MW electricity target conditional on successful confirmation. Its direct-use roadmap places market-and-resource assessment in progress, pilot demonstrations as planned, and commercial expansion in the future. These labels should be retained when discussing investment readiness. Source: https://www.tgdc.go.tz/projects

The implication is not to postpone market work until all drilling is complete. Demand mapping and resource investigation should inform one another. A promising industrial customer can sharpen the required temperature and delivery specification; resource evidence can rule out unsuitable applications before expensive detailed design.

However, infrastructure commitments should follow the weakest essential evidence. If the resource is unconfirmed, a credible buyer cannot remove geological risk. If there is no viable customer, a successful well cannot remove demand risk. Staged appraisal should expose both rather than combine them into one reassuring readiness label.

Uganda: regulatory recognition is a starting point

Uganda's official mining-cadastre guidance lists a licence for extraction of geothermal resources for direct use. That is evidence that the activity has a defined licensing category, not confirmation that a particular applicant has land, environmental, water, construction or commercial permissions. Source: https://dgsmportal.minerals.go.ug/site/CustomHtml.aspx?PageID=ddaeb30c-cdba-4d8d-8052-baeaf415f002&guidAsset=f4e49f48-2812-4f05-a408-c8dfd7b8c30c

The regional partnership can transfer appraisal methods, technical experience and lessons from failed assumptions. It cannot transfer Kenyan reservoir performance or customer economics. Uganda's first screening task should join the resource investigation to a geographically specific market assessment and a verified legal pathway.

This comparison deliberately avoids a national league table. The unit of analysis is the resource-customer pair. A well-prepared site in an earlier-stage market may be more credible than a poorly matched industrial park in a mature one.

STONECOMMS ORIGINAL SYNTHESIS: THE HEAT-SERVICE READINESS TEST

• Assess the resource and the customer together, but retain separate evidence for each.
• Contract for useful heat at a defined delivery point and service standard.
• Expand only when operating evidence supports the next commitment.

StoneComms original synthesis: The following five-part test combines IRENA's distinction between heat and electricity, the regional market-readiness programme, TGDC's staged roadmap and GRMF HEAT's requirement to assess both resource and direct-use markets. The analytical bridge is that an industrial heat project succeeds only where these independent conditions overlap. This is a proposed diligence framework, not a statistically validated predictor of project success.

1. Resource fit

Define the evidence supporting deliverable temperature, flow and resource life. Identify uncertainty, necessary testing, reinjection requirements and operating limits. Record which assumptions depend on future drilling and who bears the loss if confirmation fails.

2. Process fit

Measure the customer's temperature and pressure requirements, annual useful-heat demand, daily load and seasonal interruptions. Determine whether the existing production line can accept the proposed service. Compare retrofit costs and shutdown time as well as the heat tariff.

3. Location fit

Cost the physical route from resource to customer, including rights of way, pumping, heat losses and maintenance access. Test the whole industrial location: water, logistics, workforce, waste management and proximity to customers. Do not relocate an enterprise solely to improve one energy input.

4. Contract fit

Specify the delivery point, metering, availability standard, temperature range, payment terms and planned interruptions. Allocate reservoir decline, customer closure, fuel-price changes, pipe failure and force majeure explicitly. A contract should reveal a risk its parties cannot carry rather than conceal it in a generic availability promise.

5. Public-value fit

Establish land and water implications, worker safety, community engagement, environmental monitoring and the distribution of benefits and costs. Compare public support with realistic alternatives. Require evidence for emissions and livelihood claims separately from engineering performance.

A project can pass an early screening without being ready for finance. Each positive judgement should carry its evidence date, remaining uncertainty and next verification step. Conversely, a weak result should lead to redesign, further investigation or a stop decision rather than an automatic request for a larger subsidy.

UTILISATION IS THE COMMERCIAL HINGE

A heat network incurs costs before it sells its first unit. Its economics therefore depend on how much useful heat paying customers actually take. Peak demand alone is a poor basis for sizing; a facility can have a large boiler and still use it only intermittently.

StoneComms recommends three demand cases for appraisal: a committed-load case using credible contractual volumes; a downside case reflecting weaker production or loss of a major customer; and an expansion case that remains outside the base financing until verified. These are scenario categories, not forecasts for any country.

Customer diversification can reduce reliance on one firm but introduces coordination cost. Small customers may need aggregation, standard connection equipment and payment services. One anchor customer may simplify the initial project while creating concentration risk. The decision should compare those costs explicitly rather than assuming that more tenants always improve resilience.

The economic comparison should use useful heat delivered to the process. An electricity tariff, a fuel price and a geothermal heat price cannot be compared directly without conversion efficiency, operating cost and service reliability. The same discipline applies to carbon accounting: compare equivalent useful service, with transparent boundaries and baseline assumptions.

FINANCE PREPARATION BEFORE FINANCING EXPANSION

GRMF HEAT explicitly supports surface studies and feasibility work that includes the available direct-use market. Its published limits include up to 80 per cent of approved eligible surface-study costs, excluding infrastructure, and up to 20 per cent of approved eligible exploration-infrastructure costs. Funding is competitive and conditional; these limits are neither guaranteed awards nor a complete construction-finance package. Source: https://grmf-eastafrica.org/grmfheat/

The important design feature is the pairing of resource and market work. Public preparation money should buy reusable evidence: tested resource assumptions, customer measurements, environmental baselines, comparative options and a credible commercial structure. A glossy prospectus without those outputs has limited investment value.

Construction finance should follow verified readiness. Grant support may be justified for shared information or early uncertainty that private capital cannot efficiently carry. Debt and equity for pipes and customer connections need a different case, supported by delivery performance, payment capacity and realistic downside protection. Public guarantees should not compensate for an absent customer or an unsuitable resource.

A practical first project may therefore be narrower than a full park. A defined customer connection can test metering, maintenance, billing and process integration. Its results can inform expansion, provided the first project is commercially meaningful and not a permanently subsidised demonstration presented as a replicable business.

ENVIRONMENTAL AND COMMUNITY CONDITIONS

Geothermal heat should not be presented as impact-free. Site appraisal must examine fluid handling, reinjection, emissions where relevant, occupational exposure, noise, land access and interaction with water users and ecosystems. The appropriate measures depend on the resource and technology; this paper has not conducted a site assessment.

For communities, the relevant questions are tangible: access to land and livelihoods, local procurement, employment quality, exposure to construction and operational risks, and the accessibility of grievance procedures. Locally available heat does not automatically deliver affordable food or better employment. Benefits must be traced through the businesses and households affected.

Monitoring should distinguish four records: physical heat delivered; business performance; environmental performance; and distributional outcomes. A successful heat meter cannot establish household welfare. A growing factory cannot by itself establish that geothermal caused the growth. Evaluation should retain competing explanations and avoid double-counting benefits across a wider industrial-development programme.

IMPLICATIONS AND COMMISSIONABLE RESEARCH

Governments should require a joint resource-and-market dossier before approving a heat cluster, while keeping geological and demand risk visible separately. Regulators should clarify the applicable permission path and the responsibilities of resource owner, heat-network operator and industrial customer. These are recommendations, not claims that current law is uniform across the three countries.

Developers should prioritise measured industrial demand and compare direct heat with the customer's actual alternatives. Financiers should ask what evidence would make them stop, not simply what support would allow the project to proceed. Industrial firms should evaluate service reliability and conversion costs alongside headline energy savings.

The next commissionable research programme has four workstreams. First, audit heat demand at a defined group of firms, including process temperatures, load curves and alternative-fuel costs. Second, commission independent resource and route diligence at candidate sites. Third, test customer contracts, payment capacity and downside scenarios. Fourth, conduct community and environmental baseline work with a monitoring plan that can distinguish potential benefits from observed change.

Regional scalability should mean transferring these methods, training and contract principles. It should not mean copying park size, tariff or resource assumptions from one country to another. The strongest institutional legacy would be a transparent set of projects advanced, redesigned and rejected on comparable evidence.

COMMUNITIES AND THE UN SDGS

SDG 7 is primary because the research concerns reliable renewable heat service. SDG 9 is directly relevant to industrial infrastructure and process integration; SDG 13 concerns the climate rationale and its verification. These are relevance classifications, not evidence that a proposed project has achieved those goals.

The all-17 assessment gives the greatest prominence to those three goals. Goals 2, 8 and 12 have conditional relationships through food processing, employment and resource efficiency. Goals 6, 10, 11 and 15 require attention as water, distribution, location and ecosystem trade-offs. Goals 1, 3, 4, 5, 14, 16 and 17 are not direct classifications of this paper; their outcomes would need project-specific evidence.

Communities and workers have not been interviewed for this desk research. Claims about incomes, safety, employment or inclusion would require direct enquiry and operating evidence. The heat-service readiness test and environmental section identify the principal verification requirements. SDG relevance and relative prominence are StoneComms interpretations of issues and outcomes documented in this research. They are not formal impact measurements, and they do not imply endorsement by the United Nations.

CONCLUSION

East Africa has a credible reason to investigate geothermal heat more seriously. Regional cooperation, preparation finance and Kenya's experience provide a starting point. They do not remove the need to establish a market at the temperature, location and reliability the resource can deliver.

The immediate opportunity is to improve the quality of commitment. Measure demand before sizing infrastructure, test the resource before promising service, and contract for useful heat before presenting a cluster as investable. That approach may produce fewer early announcements, but a more defensible pipeline of industrial energy projects.