Last Updated 3 hours ago by Kenya Engineer

Political speeches describe destinations. Engineering has to describe the system that gets there.

That difference is visible in Ndindi Nyoro’s statement, “What We Must Achieve for All Kenyans.” It identifies real national pressures: expensive and unreliable production inputs, weak job creation, food insecurity, inadequate health capacity, uneven infrastructure, thin industrialisation and fragmented public systems. It also recognises something politicians do not always say aloud—that governance is harder than campaigning and that implementation depends on prior thought.

The statement therefore deserves engagement on its substance. But a list of outcomes is not yet a programme. A programme needs baselines, designs, demand forecasts, institutions, land and water, procurement routes, capital and operating budgets, risk allocation, maintenance plans and acceptance tests. It must also disclose opportunity cost. “No additional taxes” does not mean “no additional burden” if the money comes from pensions, county allocations, foregone dividends, tax concessions, borrowing or another public service.

The most useful contribution Kenya’s engineering community can make is to translate the headline proposals into questions that can be costed and checked.

The engineering agenda inside the speech

The statement contains at least seven matters with a substantial engineering dimension:

lowering electricity costs and expanding generation for manufacturing;
using a KenGen stake to raise capital and changing the role of independent power producers;
developing coastal special economic zones;
expanding irrigation, value addition, post-harvest infrastructure and rural markets;
opening northern Kenya to renewable-powered data centres;
scaling tourism and health facilities; and
creating a state-controlled company for health and e-government systems.

It also proposes that every Cabinet Secretary receive a job-creation target. That sounds administrative, but it affects how infrastructure is selected and reported. A road, data centre or irrigation scheme can create many temporary construction jobs while supporting relatively few permanent ones. Without common definitions, ministries will be rewarded for counting activity rather than durable employment.

The table below gives the short version of the engineering test. The detailed discussion follows.

Proposal	What already exists	Main engineering or delivery gap	Stronger formulation
Fully liberalise power; halt IPPs	Generation is already liberalised; EPRA has an open-access framework and Kenya Power buys from KenGen, IPPs and neighbouring systems	Market design, grid capacity, loss reduction, transparent procurement and system flexibility are not defined	Competitive auctions and open-access rules tied to the least-cost plan, published assumptions and grid-readiness milestones
Raise KSh150bn from 35% of KenGen	KenGen is listed; Government owns 70%	The stated value is far above the current market value of a 35% block; sale structure and loss of control/dividends are unclear	Independent valuation, explicit transaction route and comparison with bonds, rights issue, project SPVs and asset recycling
Make large parts of coastal counties SEZs	Dongo Kundu SEZ already covers about 3,000 acres; Lamu port-linked SEZ development is planned	A broad label does not supply land, power, water, wastewater, logistics, tenants or customs control	A few serviced, industry-specific zones with published utility capacity, tenant pipeline, export and employment milestones
Open northern Kenya through irrigation	The National Irrigation Sector Investment Plan already targets ASAL irrigation and 1.5m new acres nationally by 2030	Water balance, soils, salinity, drainage, tenure, markets and operating finance are absent	Basin-level feasibility followed by staged, water-efficient schemes measured in value and nutrition per cubic metre
Build green data centres from sun and wind	Konza’s Tier III national facility is expanding; an Olkaria green data-centre campus and KenGen storage pilots already exist	Solar and wind are variable; data centres need firm power, storage, redundant fibre, cooling, security and customers	Competitive site studies using power, fibre, cooling water, latency, skills, demand and disaster-recovery criteria
Grow to 10m international tourists	Kenya recorded about 2.7m international arrivals in 2025	Airport and destination capacity, water, waste, conservation limits and job arithmetic are not shown	A staged capacity plan linking arrivals to spend, nights, regional spread, verified jobs and infrastructure service levels
Create KDSS for health and eCitizen	The Digital Health Act already created the Digital Health Agency; health-information exchange and eClaims are being implemented	Duplication, concentration risk, data governance and the rationale for a 49% public float are unresolved	Open digital public infrastructure with clear agency ownership, interoperable standards, independent cyber assurance and separable operators

1. Cheaper power requires a system plan, not one lever

Nyoro is right on the central relationship: Kenya will struggle to deepen manufacturing while electricity is costly, volatile or unavailable when production needs it. He is also right that adequacy matters as much as price. An industrial customer loses money not only through the tariff but through interruptions, voltage problems, backup generation, damaged equipment and idle labour.

The difficulty is that the proposed remedy compresses several different reforms into “liberalise the power sector completely.” Kenya is not starting from a state generation monopoly. EPRA describes electricity generation as already liberalised, with licensed public and private producers. The sector is unbundled across generation, transmission, distribution and regulation, even though Kenya Power remains the dominant retailer and bulk buyer. EPRA now publishes electricity-market, bulk-supply and open-access regulations for 2026.

The useful policy question is therefore: which remaining monopoly functions should be opened, under what market rules and with what protections for system reliability and stranded costs?

Open access can allow a large consumer to buy from a generator while paying a transparent charge for use of the network. Competitive procurement can expose the price of new capacity. Time-of-use tariffs can move flexible industrial load away from the peak. None of these removes the need for a system operator, reserves, frequency control, transmission reinforcement, metering, settlement, credit security and a supplier of last resort.

Kenya Power’s 2025 integrated report shows why generation price alone is an incomplete target. System losses improved but remained 21.21% in the year to June 2025. Power-purchase costs were KSh144.7 billion. The utility’s approved 2024–2043 Least Cost Power Development Plan identifies grid reinforcement, storage, imports, geothermal, selected peaking capacity and carefully timed variable renewables. It also identifies a temporary adequacy gap and says generation investment will come through both KenGen and IPPs.

A serious low-cost-power programme would publish separate targets for:

average generation purchase cost and its foreign-currency exposure;
technical and commercial losses by feeder;
outage frequency and duration for industrial customers;
reserve and flexibility requirements as wind and solar grow;
transmission constraints and renewable curtailment;
the share of new capacity procured competitively; and
the total delivered cost, including network, taxes and pass-through charges.

That is less dramatic than “complete liberalisation,” but it is the route by which consumers can tell whether reform lowered a real bill without weakening the grid.

2. A blanket halt to IPPs would remove capital as well as contracts

The call to “halt IPPs unless they are publicly done” needs definition. If it means no more opaque, bilaterally negotiated deals, the concern is reasonable. Kenya should competitively procure capacity, disclose evaluation criteria, publish non-commercial PPA terms, stress-test foreign-exchange and demand risk, and make the cost of guarantees visible.

If it means ending privately financed generation, it conflicts with the current least-cost plan and removes a major source of project capital and risk-bearing. IPP is an ownership and financing structure, not a fuel and not a synonym for an expensive plant. An IPP may own geothermal, wind, solar, hydro or thermal capacity. The price depends on resource quality, financing, procurement, construction risk, contract design, currency, capacity payments and dispatch—not on the three letters alone.

Kenya Power says its medium-term plan includes acceleration of near-term IPPs and that 57 projects covering small hydro, geothermal, wind and solar with storage were under negotiation. A sudden blanket stop would either cancel capacity, shift its financing to already constrained public balance sheets or delay projects while a replacement structure is created.

The better alternative is a transparent auction programme tied to the least-cost plan and to confirmed grid capacity. Bidders should compete on an evaluated tariff that includes connection, storage or flexibility obligations where necessary. Bid bonds and development milestones can discourage speculative projects. Independent review should examine whether demand exists before a long-term capacity obligation is signed.

Kenya Engineer made a similar point in its earlier examination of energy auctions: competition can improve procurement, but an auction is not a substitute for demand planning, oversight or grid readiness.

3. The KenGen valuation does not yet add up

The speech proposes raising “close to KSh150 billion” from a 35% KenGen stake and reinvesting the money in generation. The objective—mobilising capital from a productive state asset rather than raising a new tax—deserves examination. The stated valuation, however, needs a public calculation.

KenGen has 6,594,522,339 issued ordinary shares. The Nairobi Securities Exchange daily price list for 17 August 2026 recorded KenGen at KSh11.30 per share. That gives a market capitalisation of roughly KSh74.5 billion. At that snapshot price, 35% of the company is worth about KSh26.1 billion before transaction costs and any block premium or discount.

To raise KSh150 billion by selling 35% of the company, investors would have to value the whole company at about KSh428.6 billion, equivalent to almost KSh65 per share—about 5.7 times the 17 August market price. A strategic investor may value assets differently from the stock market, and a proper valuation would examine cash flows, replacement cost, debt, geothermal resources and future projects. But “the right valuation” cannot simply be asserted when a public market already provides a reference.

The transaction structure is equally important. Government currently owns 70%. Selling 35 percentage points of the company—half of the Government’s present holding—would reduce the state to 35%, surrendering majority ownership and a share of future dividends. Issuing new shares to raise money would produce a different dilution. Pledging assets, borrowing against cash flows or using project companies would create different risks again.

Before any sale, the public should see a comparison of at least five options:

a secondary sale of existing government shares;
a primary rights issue in which new money enters KenGen;
a green or infrastructure bond backed by company cash flows;
project-specific equity or debt in ring-fenced generation assets; and
asset recycling, where a mature asset funds a new one without transferring the entire company’s control.

Each option should disclose capital raised, cost of capital, currency, guarantees, dividend sacrificed, control rights and the projects that are construction-ready. Sale proceeds are one-off. New generating assets also require transmission, system services, maintenance and customers able to buy the energy.

4. Coastal SEZs need serviced boundaries, not county-sized promises

The idea of using the coast as an export-oriented manufacturing platform is not new, and that is not a criticism. Good national projects often need continuity across governments. Dongo Kundu is already a public SEZ of about 3,000 acres next to the Port of Mombasa, the Southern Bypass, the SGR and Moi International Airport. Kenya Ports Authority has also identified Lamu port-linked SEZ development.

The engineering problem is not the absence of a label. It is delivery of a bankable industrial location.

The Special Economic Zones Act requires a defined perimeter and considers land title, topography, access, water, power, sewerage, telecommunications, solid and liquid waste, fire safety, security, medical facilities and environmental requirements. Declaring “most parts” of Mombasa, Kwale, Lamu and other coastal counties as SEZs would make customs administration harder and extend tax preferences over a vast area without guaranteeing a single serviced plot. It would also sacrifice public revenue before the additional investment is known.

Shenzhen is a useful ambition but a poor shortcut. Its result came from ports, large-scale urban and industrial infrastructure, municipal capability, supply-chain depth, labour and housing, national reform and decades of investment. A Kenyan zone will not reproduce that outcome by changing the tax address of land.

A stronger proposal would select a small number of coastal nodes based on anchor demand. Dongo Kundu could report available megawatts, water and wastewater capacity, road and berth interfaces, titled industrial plots, signed tenants, local-supplier opportunities and export value. Lamu should be sequenced against actual port traffic, security, environmental and community obligations, and the cost of connecting utilities. Candidate sectors should be chosen by logistics and market evidence—such as agro-processing and cold chains, assembly, maritime services or construction materials—not by a generic manufacturing wish list.

5. Irrigation is already policy; performance is the missing argument

The speech proposes irrigation and wider economic “opening” of northern Kenya, use of idle land, better seeds and fertiliser, model farms, crop selection, nearby value chains and more processing. Much of this direction is consistent with existing policy.

Kenya’s National Irrigation Sector Investment Plan aims to develop 1.5 million acres by 2030 and says only 21% of national irrigation potential is currently used. One of its five pathways is specifically for ASAL areas, using small and medium water storage, spate irrigation, groundwater and managed aquifer recharge. The proposal is therefore not entering an empty field. It needs to say what will be done differently from NISIP and how failed or underperforming schemes will be corrected.

The phrase “idle land” is especially risky. Land that is not cultivated may be grazing territory, a wildlife corridor, a seasonal floodplain, an aquifer-recharge area or community land with unresolved rights. The limiting resource in much of northern Kenya is water, not acreage.

Every proposed irrigation scheme should pass, at minimum, these tests:

dependable surface- and groundwater yield under drought and climate variability;
environmental flow and downstream-user requirements;
soil depth, salinity, sodicity and drainage;
land tenure, pastoral mobility and community consent;
pumping energy and spare-parts supply;
crop water requirement and gross margin per cubic metre;
storage, cold chain, processing and contracted market;
who owns and maintains the intake, canals, pumps and meters; and
a tariff or financing plan that covers operations and renewal.

The speech’s emphasis on edible-oil crops addresses a real dependency: KIPPRA reports that Kenya produces only about 34% of its edible-oil requirement. But the claim that edible oils are the country’s “second-highest import by value” is too broad without a year and commodity classification. It may refer to food imports rather than all national imports. Petroleum, machinery and other industrial goods dominate different trade classifications. Policy should use a precise customs code and baseline so that success can be measured.

Crop selection should also optimise more than the highest farm-gate price. Engineers, agronomists and economists need a combined measure of water productivity, climate risk, transport, processing utilisation, nutrition, market depth and farmer margin. A high-value crop with no cold chain or buyer can destroy more value than a lower-priced crop with a dependable market.

Using schools as periodic rural markets may make use of existing access and buildings, but it is not costless. Water, toilets, drainage, solid waste, traffic separation, food safety, storage, fire access, cleaning and protection of children and school assets would have to be designed and funded.

6. A green data centre is a reliability project before it is a solar project

Northern Kenya has excellent solar and wind resources. That makes green data centres plausible, not automatic.

A data centre consumes power continuously, while the sun sets and wind output changes. A credible facility needs firm supply, storage or another backup path, power-quality equipment and maintenance support. It also needs at least two physically diverse high-capacity fibre routes, low enough latency for the target customers, cyber and physical security, a skilled operations team, fire suppression and a cooling design suited to a hot and dusty environment. Water use becomes an important site question if evaporative cooling is proposed.

Kenya already has relevant projects against which a northern proposal should compete. Konza’s National Data Centre is Tier III certified and is being expanded as a sovereign cloud platform. KenGen has demonstrated battery storage for a modular data centre and is developing a geothermal-powered green-energy park at Olkaria. Geothermal’s firm output is a different proposition from a solar-and-wind-only site.

The engineering answer is a competitive site study, not a predetermined county. Candidate locations should be scored on delivered firm-power cost, renewable additionality, fibre diversity, latency, cooling energy and water, physical risk, skills, customer demand and disaster-recovery value. The public business case should include power usage effectiveness, water usage effectiveness, carbon intensity and an uptime target.

A northern facility may prove valuable as an edge or disaster-recovery location and as an anchor for fibre and energy investment. It should be built because customers and resilience justify it, not because a map shows abundant sunshine.

7. Ten million tourists is an infrastructure scenario, not a comparison of country size

Kenya recorded about 2.7 million international arrivals in 2025, up from roughly 2.47 million in 2024. A target of 10 million is therefore about 3.7 times the present flow. It may be an ambition worth testing, but comparing Kenya’s physical size and attractions with Morocco or Albania does not demonstrate capacity or demand.

Tourist flows depend on air seats and fares, source-market income, entry friction, safety, brand, accommodation, product, regional land borders, length of stay and how arrivals are counted. More visitors are not necessarily more value if they stay briefly, concentrate in overloaded destinations or reduce the environmental quality on which tourism depends.

The associated jobs claim also needs a model. The speech projects three million direct jobs and then applies three indirect jobs to every direct job. Multipliers cannot be selected as a universal rule. They must come from a current Tourism Satellite Account or input–output model that avoids double counting and distinguishes direct, indirect and induced employment. Kenya’s earlier Tourism Satellite Account estimated about 990,000 direct jobs in 2019; the current baseline should be updated before a new figure is announced.

For engineers, a 10-million-arrival scenario means testing:

runway, terminal, air-navigation and border-processing capacity;
roads, rail, ferries and last-mile access to destinations;
hotel and community water supply, wastewater and solid waste;
electricity, digital connectivity and emergency services;
park, beach, marine and heritage-site carrying capacity;
workforce housing and municipal services; and
climate resilience for coastal, wildlife and mountain assets.

A more defensible target would be staged and multidimensional: international arrivals, domestic trips, average spend, length of stay, geographic spread, repeat visits, verified jobs and environmental service levels. Marketing can fill capacity; it cannot substitute for it.

8. The proposed government systems company risks duplicating work already under way

The statement proposes Kenya Data and Systems Services, majority-owned by government with 49% sold to the public, to manage the health system and eCitizen. The attraction is understandable: critical public platforms should not depend on opaque arrangements or inaccessible vendor systems.

But the institutional baseline is missing. The Digital Health Act, 2023 already established the Digital Health Agency and charged it with an integrated health-information system. Health-information exchange, biometric identification, facility onboarding, interoperability standards and electronic claims are already being implemented. In March 2026, the Ministry of Health reported that 10,277 facilities had been onboarded. eCitizen also sits within an existing digital-government governance structure.

Creating a new company could consolidate technical capacity. It could equally duplicate the Digital Health Agency, concentrate two critical national platforms in one failure domain and blur the line between regulator, system owner and commercial operator.

Majority government ownership does not, by itself, guarantee security, interoperability or public control. Nor does listing 49% solve procurement or accountability. A listed company has a duty to shareholders and will seek revenue; a core public-data platform has duties of legality, continuity, privacy and equal access. Those incentives can coexist only if the service obligations and revenue model are explicit.

A better digital architecture would separate five roles:

Parliament and regulators set rights, standards and accountability.
Each service agency remains responsible for its statutory service and data.
Shared digital public infrastructure provides identity, payments, notifications and interoperable exchange through open standards.
Certified public or private operators compete to host and support separable components under measurable service levels.
Independent teams conduct security testing, privacy impact assessments, uptime audits and disaster-recovery exercises.

Health data needs particularly strict segmentation, consent and access controls, audit logs, clinical safety, retention rules and tested portability. The objective should be that no ministry is trapped by a supplier and no single failure disables both health and general government services. Corporate form is secondary to architecture and law.

9. Health infrastructure cannot be separated from pension trade-offs

The speech proposes halving NSSF contributions and redirecting roughly KSh40 billion to healthcare, including chronic care, facility upgrades and supplies. The arithmetic has a visible basis: NSSF reported member contributions of about KSh84 billion for 2024/25, so half is of the stated order.

The money is not unallocated revenue. It is workers’ and employers’ pension saving governed by the NSSF Act. Diverting it would reduce funded retirement assets and future investment income, require legal and actuarial changes, and transfer a cost from current health financing to future retirees. The proposal may still be argued as a social choice, but it cannot be described as burden-free.

For the engineering portion—facility upgrades—the first step should be a national asset-condition and service-capacity audit. A hospital is a system of power, water, medical gases, ventilation, infection control, waste, sterilisation, laboratories, cold chain, digital networks and maintainable equipment. Procuring machines without rooms, consumables, technicians, calibration and service contracts produces stranded assets.

Funding should therefore be tied to facility service plans and lifecycle budgets, not only construction or equipment counts. Biomedical, electrical, mechanical, civil, water and ICT engineers should be embedded in planning and acceptance. The proposed healthcare-workers commission addresses staffing, but it does not replace technical asset management.

10. Jobs must be measured in a way ministries cannot game

Requiring every Cabinet Secretary to report net jobs could improve focus. It could also generate impressive but meaningless totals unless KNBS owns a common method.

The speech itself illustrates the danger. It cites informal employment at 83.8% and then says most of those workers are in small-scale agriculture. KNBS’s employment table explicitly says its informal-sector stock excludes small-scale farming and pastoralist activities. The 83.8% statistic may describe the formal–informal split in the measured non-farm employment series; it cannot support the agricultural conclusion made from it.

A national jobs dashboard should distinguish:

gross jobs announced from net jobs after closures and displacement;
persons from full-time-equivalent jobs;
construction-months from permanent operations jobs;
formal payroll jobs from self-employment estimates;
direct project employment from indirect and induced effects;
Kenyan labour from total labour; and
job count from earnings, productivity, safety and duration.

The reporting agency should publish methods and revisions and prevent several ministries from claiming the same job. Infrastructure should be judged by the durable economic activity it enables, not simply the labour present during construction.

The sacrifices that should be stated plainly

The speech promises that several measures can be funded without additional taxes. That is possible only in the narrow sense that funds can be reassigned. Every major proposal still has a trade-off.

Proposed gain	Likely sacrifice or risk if poorly designed
Lower electricity price	Utility revenue, network investment or reliability if tariffs fall without cost reduction
KenGen sale proceeds	State control, future dividends and exposure to an undervalued one-off sale
Halted IPPs	Private capital, project speed and risk transfer; more pressure on public borrowing
County-scale SEZ incentives	Tax revenue, customs simplicity, land rights and environmental services
Redirected NSSF contributions	Retirement savings, compounding investment returns and legal certainty
Rapid irrigation expansion	Water, ecosystems, pastoral mobility and unaffordable operations
Northern data-centre location	Higher cooling, fibre and redundancy cost if resource maps override demand and reliability
Ten million tourists	Congested airports and destinations, water stress, waste and conservation damage
One government systems company	Concentration of cyber, operational and governance risk

Kenya’s fiscal position makes this candour more important. The National Treasury’s FY2026/27 Annual Borrowing Plan estimates gross financing needs of about KSh1.996 trillion, including KSh1.145 trillion for the deficit and KSh851 billion for maturing debt. Engineering programmes will compete for scarce capital and for the technicians and institutions that operate it. Sequencing is not pessimism; it is delivery.

What an engineering-ready national programme would contain

The statement is strongest when it links energy, production, value addition, infrastructure and jobs. Those sectors do reinforce one another. It is weakest where it assumes that a target, institution or financing source creates capacity merely by being named.

An engineering-ready version should attach the same one-page project test to every promise:

Baseline: What is the verified present capacity, cost and service level?
Demand: Who needs the output, when, where and at what price?
Options: What non-build, rehabilitation, market and new-build alternatives were compared?
Design basis: Which codes, loads, resource data and climate scenarios govern?
Capital and lifecycle cost: What are the capex, opex, renewals and decommissioning obligations?
Funding and risk: Who pays, borrows, guarantees and carries demand, currency and construction risk?
Land, water and environment: Which rights, permits, communities and cumulative impacts apply?
Delivery: Who procures, designs, checks, builds, accepts, operates and maintains?
Milestones: What can the public verify each quarter before more money is released?
Outcome: Which independent measure proves lower cost, higher reliability, more output or durable jobs?

That framework would also improve the proposal for monthly or quarterly reporting to the President. Reports should come from open project data and independent verification, not only from the ministry responsible for the promise.

An invitation to make the politics more technical

Nyoro’s statement has put several engineering questions into the political arena, and that is welcome. Kenya needs political leaders to discuss power-market design, industrial utilities, irrigation performance, public digital infrastructure, asset finance and the physical capacity behind tourism and health.

The proper response is neither applause nor dismissal. It is to ask for the calculation.

What does “fully liberalised” mean at the meter and at the control room? What transaction produces KSh150 billion from KenGen without selling a national asset too cheaply? Which coastal plots will have power, water, wastewater and tenants? Which northern water source can sustain which crop through a dry cycle? Which fibre routes and firm power make a data centre bankable? Which airport, water and conservation investments precede ten million visitors? What problem remains for a new systems company after the Digital Health Agency’s work is counted?

Answering those questions would not make the agenda less ambitious. It would turn ambition into a design that can be debated, priced, built and audited. That is where Kenya’s engineers should enter the conversation.