Powering Kenya

The Paradox of Kenyan Power

Why does Kenya keep building generating capacity while importing electricity?

Olkaria I returns to a system built on four facts that seem to contradict each other. The country reports almost 4,000 megawatts of generating capacity against a peak demand of about 2,500. It bought one in every eight units it used last year from Ethiopia and Uganda. It switched off, or curtailed, 404 gigawatt-hours of its own geothermal and wind output. And the average customer spent more time without power in FY2026 than in FY2025.^[1]

If Kenya had too much power, it would not need imports. If it had too little, it would not throw output away. The evidence supports neither. Four questions resolve most of the contradiction: which plants does it count, at what time of day, at which point on the network, and who pays for it?

All figures are for the year to 30 June 2026 and come from the regulator's statistics report.^[1] Hours without power are EPRA's mean monthly SAIDI, compared on the same basis as the FY2025 report.^[2]

What the 3,987 MW headline counts

EPRA's headline of 3,987.2 MW at June 2026 counts everything: plants that factories and farms run for themselves (captive capacity, 676.6 MW), small isolated grids, and a 200 MW import contract with Ethiopia that is not Kenyan machinery at all. Remove those and the domestic grid-connected nameplate is 3,063 MW.^[1] Even that is a rating stamped on equipment, not a promise of output. Hydro depends on rain, wind and solar on weather and hour, geothermal on steam pressure, and every plant breaks down sometimes. The figure that matters for security, the capacity available when demand peaks around 7 to 9pm, is not published.

Households, not factories, drove demand growth

Served peak demand rose from 2,177 MW in FY2024 to 2,316 MW in FY2025 and 2,514 MW in FY2026, growth of 6.4% and then 8.6%. KenGen reports a further record of 2,549 MW on 15 July 2026.^{[1] [7]} Consumption billed across tariff categories rose 9.8% to 12,445 GWh. But households, not industry, supplied 61.6% of that increase; large commercial and industrial use grew 5.3%.^[1] That could reflect new connections, better billing, hotter weather or suppressed demand returning. It is not, on its own, evidence of an industrial boom (Chapter 5).

Curtailment is a night-time problem

Geothermal curtailment is concentrated between midnight and 5am, when demand is lowest and geothermal plants, which cannot easily ramp down, produce more than the grid needs. The peak comes in the evening. Annual totals cannot tell us whether an imported unit displaced a usable Kenyan unit in the same hour and place. Curtailment fell 40% in FY2026, which EPRA attributes to higher night-time demand.^[1]

How investment becomes shareholder value

Electricity investment creates value only if it passes through a chain of handoffs, and each one can leak: money spent → power delivered when needed → bills paid → cash left after reinvestment and debt → a return to a shareholder who bought at today's price. The chapters that follow test each link. Our conclusion, in brief: an uneven investment phase is clearly visible in selected generation, network and trade projects, but a synchronised, fully financed national cycle is not established, and the evidence does not yet show a robust winner among the listed shares.

Electricity 101

How do generation, transmission, distribution and pricing work?

How electricity reaches a customer

A power station turns steam, falling water, wind, sunlight or fuel into electricity. A transformer raises the voltage so it can travel long distances with less loss, along the 132, 220 and 400 kV lines of the transmission grid, or the high-voltage DC line from Ethiopia. Substations step the voltage down again into the distribution network of poles and cables that reaches homes and businesses, where a meter records what each customer uses. Every stage loses some energy as heat; some is also lost to theft and faulty meters. Kenya Power bought 14,472 GWh in FY2025 and could not bill 3,069 GWh of it, 21.2%.^[3]

The evening peak

The grid has no meaningful storage, so supply must match demand second by second. The system operator does this by deciding which plants run, a process called dispatch. Kenyan demand is lowest after midnight and highest between about 7 and 9pm, when lights, cooking and television overlap. Across FY2025 the average load was about 1,652 MW against a peak of 2,316 MW: the grid ran at roughly 71% of its peak on average, a figure engineers call the load factor.^[24] The gap between the average and the peak is why a country can have "spare" capacity at 3am and still struggle at 8pm.

Plants play different roles in that daily cycle. Geothermal runs flat out day and night and is hard to turn down, which is why it is curtailed in the early hours. Hydro can ramp up in minutes and stores water in dams, so it follows demand. Wind and solar produce when the weather allows. Diesel and heavy-fuel plants are expensive per unit but can be started for the evening peak or a dry year.

Five kinds of capacity

Only the first two of these measures are published.

Definitions follow the research's measurement boundaries; published values from EPRA 2026, Table 2.1.^[1]

How a unit of electricity is priced

Generators sell to Kenya Power under long-term power purchase agreements. Most split the price into a capacity payment, which covers the fixed cost of building and maintaining the plant and is paid if it is available, and an energy payment for each unit delivered, which covers fuel and running costs. Kenya Power then sells to customers at a regulated tariff: a base energy charge, plus monthly pass-through charges for fuel, exchange rates and inflation, plus levies and VAT.^[14] The same kilowatt-hour therefore has several "costs", depending on where you stand.

Hydro at KES 3.07 comes mostly from old, fully paid-for dams. Thermal at KES 34.50 includes off-grid diesel stations and capacity charges on plants that run few hours. The chart is not a ranking of the cheapest technology, because it mixes old assets with new. It does show that Kenya Power's total purchase bill was KES 10.72 per unit it bought, but KES 13.60 per unit it sold, because about one unit in five was lost between purchase and bill.^[3]

Anatomy of Kenya's Electricity Market

Who are the major players, and how do their contracts and incentives interact?

Kenya runs a regulated, contract-based system, not a market where generators sell to the highest bidder. KenGen and the independent power producers (IPPs) sell almost all their output to Kenya Power under power purchase agreements approved by the regulator, EPRA. Kenya Power then distributes it and bills customers at tariffs EPRA also approves.^[3] Two state companies sit upstream and alongside: the Geothermal Development Company (GDC), which drills and sells steam, and the Kenya Electricity Transmission Company (KETRACO), which builds high-voltage lines.

Kenya Power owns a third of the transmission grid

At June 2026 Kenya had about 10,100 circuit-kilometres of transmission line: 6,399 owned by KETRACO and 3,702 by Kenya Power.^[1] So Kenya Power is more than a distributor, and not every line loss belongs to it. Gazette Notice 155, published on 14 January 2022, designated KETRACO as the national system operator, the body that decides which plant runs when.^[23] KETRACO's own audit follow-up still described the handover from Kenya Power's national control centre as outstanding, and a new control-centre contract announced in January 2025 had a 36-month delivery plan. We found no certificate showing the transfer was complete by our cutoff.

What an investor can own

The National Treasury owns about 70% of KenGen (69.99% at June 2025) and the government 50.1% of Kenya Power.^{[6] [3]} KETRACO and GDC are wholly state-owned. The IPPs are separate project companies with their own shareholders; none trades on the NSE. OrPower4, the Olkaria geothermal IPP, belongs to the US-listed Ormat Technologies, which gives exposure to Kenya diluted by Ormat's assets elsewhere.^[16] For a Kenyan retail investor, direct ordinary-share exposure to this sector means KenGen (KEGN) or Kenya Power (KPLC). Chapter 9 returns to the alternatives.

Two EPRA approvals govern the money

Two separate EPRA decisions govern the money. The first approves each generator's power purchase agreement with Kenya Power. The second approves the retail tariff customers pay, currently the schedule in Gazette Notice 3899, effective 1 April 2023.^[14] Approval of a generator's contract does not guarantee that the retail tariff will recover its cost on time, or that customers will pay. Much of the sector's financial risk sits in that gap.

Generators are paid for being available

Most Kenyan PPAs split the price in two. A capacity payment is paid for keeping a plant available, whether or not it is dispatched; an energy payment is paid per unit delivered. For KenGen in FY2026, capacity payments were KES 32.96bn of KES 42.59bn electricity revenue: 77%.^[6] That protects the generator against low demand and moves that risk to Kenya Power and, ultimately, customers. Paying for a plant that is rarely used can be rational if it is there when the hydro dams are low or the evening peak arrives. Whether each contract is good value depends on terms that are mostly not public.

Losses above 16.5% are Kenya Power's cost

EPRA's tariff allows Kenya Power to recover the cost of a set level of system losses: 19.5% in FY2023, 17.5% in FY2025 and 16.5% in FY2026. Kenya Power reported actual losses of 21.21% in FY2025 and 18.58% in FY2026. The excess is a cost the company absorbs.^{[14] [4]} EPRA publishes a different loss series (23.36% and 21.38% for the same years); the two have not been reconciled, and we report both rather than choose.^{[1] [2]}

The 2026 tariff application was withdrawn

Kenya Power filed for a new tariff for the three years 2026/27 to 2028/29 on 31 March 2026. The Ministry withdrew it on 3 June 2026 and EPRA cancelled the public hearings in a notice published on 5 June.^[15] The 2023 schedule therefore continues, with monthly fuel, currency and inflation adjustments. Kenya Power's future margin depends on that deferred decision.

Dollar contracts put currency risk on Kenya Power

Many renewable PPAs are priced or indexed in US dollars even though they burn no imported fuel. That protects the generator and its lenders, and passes exchange-rate risk to Kenya Power and its customers. Ormat disclosed $29.5m overdue from Kenya Power at December 2025, of which $21.1m was paid in January and February 2026.^[16]

The Generation Economy

How do geothermal, hydro, wind, solar, thermal and imports compare economically?

In FY2026 Kenya's grid received 6,421 GWh from geothermal plants, 3,554 from hydro, 1,932 from wind, 1,309 from thermal plants and 471 from utility-scale solar, plus 1,914 GWh of imports. Geothermal, hydro, wind and solar together supplied 78.9% of the total.^[1] That share reflects decades of state investment at Olkaria and on the Tana River.

Value depends on when a plant can run

Geothermal is about 30% of grid capacity but 41% of energy, because it runs around the clock. Thermal is 18% of capacity and 9% of energy: it is expensive per unit, but it is there for dry years and evening peaks. Hydro can be turned up quickly and stores water, so it helps balance the grid, but its output fell to 2,569 GWh in drought-hit FY2023 from 4,141 GWh two years earlier.^[3] Wind and solar produce when the weather allows, which is rarely at 8pm for solar. A plant's value depends on when it can produce, not only on how cheaply.

What new plants cost

The government's June 2024 planning assumptions put the upfront cost of new solar at about $1.28m per MW, wind at $1.89m and geothermal at $3.50–3.81m.^[17] Globally, IRENA reports 2025 average costs for new solar of $44 per MWh, onshore wind $33 and geothermal $89.^[20] They are not Kenyan tariffs: financing costs, grid connection, currency risk and the different time profile of each technology can account for much of the gap. Geothermal's higher cost buys output that is steady day and night.

Geothermal means managing a reservoir

Kenyan technical papers describe natural pressure decline at Olkaria, delays in connecting replacement ("make-up") wells, and six reinjection wells lost to corrosion. At Menengai, GDC bears the cost of drilling and sells steam to private plant owners, so their published plant costs exclude a large public investment upstream.^[6] Two 35 MW Menengai plants were added in FY2026 and produced 116 and 20 GWh respectively.^[1] A new turbine does not remove steam risk (Chapter 6).

Infrastructure and Demand

Where are the grid's constraints, and what will drive future consumption?

Reliability worsened while losses fell

Losses improved on EPRA's measure from 23.36% to 21.38%, yet the time each customer spent without power rose from about 9.4 to 13.2 hours a month.^{[1] [2]} The data do not tell us how much of that deterioration comes from transmission, distribution, generation or changes in reporting. They do show that more generation alone would not fix it. Some network projects show results: the Sondu–Ndhiwa line gave South Nyanza a second supply route, and Mariakani's energisation in December 2025 allowed the coast corridor to run at 400 kV.^[11]

KETRACO's 2026–2045 master plan puts the investment need at about $5.3bn, with $350m secured and a $4.95bn gap, of which $1.95bn falls before 2030.^[11] These are planning envelopes, not funded orders. One private route is taking shape: an Africa50–PowerGrid transmission concession signed on 15 December 2025, with a project cost of $311.47m and a availability-based annual revenue requirement of $43.12m.^[12] That revenue is a cost to the sector, not income for an NSE-listed company, and the concession has not reached financial close.

Imports have grown twelvefold since FY2020

Imports rose from 161 GWh in FY2020 to 1,914 GWh in FY2026, up 24.8% in the last year alone. Ethiopia supplied 1,585 GWh, nearly all of it through the high-voltage DC line; Uganda 322 GWh.^[1] The Ethiopia interconnector is rated 2,000 MW, but EPRA still describes 200 MW of contracted imports. Lender documents describe a larger contractual ramp and identify voltage-support constraints on the Kenyan side; the relevant equipment package has been deferred to 2028.^{[13] [11]} Kenya Power reports a 100 MW Ethiopia-to-Tanzania wheeling trial, but the regional power pool's independent market operator was still pending in April 2026, and we found no live settlement data for a regional day-ahead market.^[4]

Who is using more electricity

From FY2016 to FY2025 Kenya Power's customer accounts more than doubled while system sales grew about 44%. Average consumption per account fell from about 1,609 to 1,128 kWh a year, because most new connections are small households.^[3] In FY2025, just 397 new large-power accounts used 110 GWh, 54% of all the energy consumed by that year's new connections. New uses are small so far: electric-mobility tariffs used 12.25 GWh in FY2026, about 0.1% of consumption, and a 2023 survey found electric cooking as the main method in 0.58% of households.^[1]

Kenya Power's own reported sales were 12,777 GWh in FY2026, which differs from EPRA's category total by 332 GWh; we treat the difference as unallocated rather than assume it is exports.^[5] And all of this is grid demand: captive solar on rooftops and factories grew 72.8 MW in a year, but how much utility energy it displaced is not measured.

Case Study: Olkaria I

Was KenGen's geothermal rehabilitation economically justified, and what could it return?

We opened this report with Olkaria I's return to service. This chapter asks whether it was worth it. Olkaria I, in Hell's Gate near Naivasha, was Africa's first geothermal plant. Its three original 15 MW units had reached the end of their lives. KenGen has replaced them with equipment rated at 21 MW each, and reported the rehabilitation complete at its AGM on 29 October 2026.^[7] The AGM release put the cost at KES 15.6bn; the annual report's project reference is KES 15.785bn, and we use that figure throughout.^[6] The engineering is public; the economics are not.

Four ways to count Olkaria I's capacity

The 2.5 MW gap between the supplier's 63 MW and EPRA's 60.51 MW is consistent with power the plant uses itself, but EPRA does not say which boundary it used.^[1] Against the old 45 MW, only 18 MW is new capacity. Against the alternative of letting the worn-out units retire, the whole output is restored. Which baseline you use changes whether KES 15.8bn looks well spent.

What is known about the deal

A screening model

Because the tariff is unknown, we built a simple whole-project model to ask what the deal would need to look like to earn its cost. Our base case assumes 63 MW gross, 4% own use, 90% availability and 90% dispatch, 0.5% annual output decline, 25 years of operation, an illustrative KES 8 per net kWh in 2027 rising 3% a year, KES 125m a year of extra steam-field maintenance and a 12% nominal hurdle rate. On those assumptions the plant produces about 429 GWh in its first full year, has a net present value of minus KES 1.10bn and an IRR of 11.1%, and needs about KES 8.41 per unit to break even.^[24] The KES 8 is our placeholder, not KenGen's tariff. A small difference in the real number flips the sign.

The tariff dominates: every KES 1 per unit moves value by about KES 2.7bn. Availability is next: at 95% the base case roughly breaks even. And the hurdle rate matters because the project is financed with very cheap foreign-currency debt (a 1% coupon on about KES 7.6bn in our model). Measured on the cash left after that debt, our base case looks positive (KES 1.45bn at a 15% hurdle), but that cheap loan carries currency risk, and a positive financing residual is not the same as cash a minority shareholder receives.^[24]

Fair values, thesis tracking and alerts when the evidence changes, for the NSE stocks you own. This research stays free to read.

KenGen and Kenya Power

How financially attractive are the two listed companies, and what are their shares worth?

The two companies sit at opposite ends of the money chain described in Chapter 3. KenGen is paid mainly for being available, under long-term contracts with one customer. Kenya Power is that customer: it buys from everyone, carries the network and collection risk, and depends on a regulated tariff. We judge both on the cash left after reinvestment and debt, not on earnings.

The business

KenGen is Kenya's largest generator, with most of the country's hydro and much of its geothermal. In FY2026 it supplied 8,975 GWh. Gross revenue rose 6.4% to KES 59.71bn; net revenue, after stripping out fuel and water costs passed straight through to Kenya Power, rose 6.6% to KES 49.51bn. Operating profit rose 4.1% to KES 14.17bn, but profit after tax fell 1.2% to KES 10.35bn as interest income on deposits shrank. The board cut the dividend from KES 0.90 to KES 0.75 a share.^[6]

Because so much income is capacity payments, a "national demand × average tariff" forecast misses how KenGen earns. More demand helps only at the margin, through energy payments and new contracts. What matters more is keeping plants available, getting paid on time and earning a return on new spending.

Earnings are not cash

KenGen reported KES 30.03bn of operating cash flow in FY2026, three times its profit, largely because depreciation is a big non-cash cost.

KenGen generated KES 27.30bn from operations and spent KES 37.50bn on capital projects, loan repayments, dividends and interest. The KES 10.2bn gap came from new borrowing and cash on hand.^[6] Borrowings were KES 107.71bn at June 2026, all of the principal in US dollars, yen or euros. Low coupons help, and some PPAs let KenGen recover currency losses when loans are repaid, but that is a timing mechanism, not an exemption from risk. Kenya Power owed KenGen KES 18.17bn in gross trade receivables.^[6]

The auditor's findings

The FY2026 accounts received a clean, "fair presentation" opinion. Separately, the auditor flagged control and lawfulness exceptions, overdue receivables, unresolved geothermal licence coordinates, and, as a key audit matter, insufficient support for applying a uniform 10% residual value to every class of depreciable asset.^[6] Lower depreciation lifts profit without adding cash. Book equity of KES 290.15bn (KES 44 a share) includes large revaluation reserves; profit was only about 3.6% of opening equity.

Valuation

At the 30 October volume-weighted average price of KES 9.90, KenGen trades at 6.3 times FY2026 earnings and about 0.22 times book value.^[8] Our discounted cash flow cases, which explicitly model replacement investment, working capital and funding through FY2035, give equity values of KES 0 (bear), KES 0.64 (base) and KES 16.82 (bull) a share. The base case also needs about KES 8.53bn of new, uncommitted borrowing to keep paying dividends at policy levels; if future payouts were suspended that need disappears, though the cash-flow value is unchanged.^[24] The gap exists because KenGen reinvests most of its cash, and our model cannot yet link about KES 17.1bn of growth spending to specific contracted projects.

The recovery

Kenya Power lost money in FY2020 and FY2023. It then earned KES 30.08bn in FY2024, KES 24.47bn in FY2025 and KES 24.99bn in FY2026. Its reported system losses fell from 23.16% in FY2024 to 21.21% and then 18.58%. The loan book fell from KES 115.41bn in FY2023 to KES 79.82bn in FY2026.^{[4] [5]}

Operating profit fell

In FY2026 revenue rose KES 18.96bn and gross profit KES 10.97bn, but operating costs rose KES 11.33bn and other income fell KES 0.54bn, so operating profit fell by KES 0.90bn to KES 38.57bn. Lower finance costs (down KES 1.64bn) turned that into profit growth of 2.1%. Operating cash flow fell from KES 39.77bn to KES 38.22bn.^[4] FY2025's profit had also been flattered by a KES 4.43bn net credit after a change in how the company estimates bad-debt provisions, against a KES 1.18bn charge the year before; without it, FY2025 earnings per share would have been about KES 10.95 rather than KES 12.54.^[3]

Net current assets turned positive at KES 1.90bn, but net receivables rose KES 20.46bn to KES 96.48bn, and postpaid debts more than 90 days old were KES 20.41bn, with large balances owed by government and rural schemes.^[4] About 78.8% of the remaining loans are in dollars or euros.

Spending needs compete with dividends

Kenya Power spent KES 28.22bn on capex in FY2026 and had KES 50.51bn of contracted commitments not yet incurred. Fully depreciated assets still in use had an original cost of KES 68.18bn: a sign of the replacement spending ahead.^[4] The Auditor-General signed an unmodified opinion on the FY2026 statements on 17 September 2026, with an emphasis on land titles and a key audit matter on revenue recognition; separate conclusions on lawfulness and controls contain exceptions, and 12 earlier matters remain unresolved.^[4]

Valuation

At KES 21.60, Kenya Power trades at 1.7 times FY2026 earnings.^[8] Its historical cash left after capex and interest (excluding interest income) was KES 9.45bn, before any debt repayment or dividend. Our repaired base cash-flow case, on moderate demand, gives KES 15.08 a share; accelerated demand gives KES 23.93 and distributed KES 12.09 (the demand cases are set out in Chapter 8). The bull case reaches KES 59.10 and the bear case zero.^[24] The single most powerful input is the cash margin we let Kenya Power keep on its revenue. We cap it at 14%; at 12% the base value is zero, at 16% it is KES 35.20. EPRA has not set any such margin, and the tariff review that might have was withdrawn.

The Next Decade

How might Kenya's electricity market evolve through 2035, and what opportunities could emerge?

Chapter 7 valued both companies on our moderate demand case. This chapter sets out the alternatives, and the policy and technology changes that could move value between customers, private developers, the state and the listed companies.

Open access and net metering

The Electricity Market, Bulk Supply and Open Access Regulations (Legal Notice 79 of 2026), gazetted on 8 May 2026, allow bilateral, spot and forward trading and give customers using at least 1 MVA on the distribution network, or 10 MVA on transmission, the right to buy from someone other than Kenya Power.^[18] EPRA approvals, network studies, contracts and wheeling charges still have to be in place. Open access could take Kenya Power's most valuable customers, or it could be slow and narrow; we do not assume either.

Under net metering (Legal Notice 104 of 2024), exported units earn a credit worth 50% of a unit, unused credits are forfeited at the end of the utility's financial year, residential systems are capped at 4 kW single-phase or 10 kW three-phase, and the first five-year programme is capped at 100 MW.^[19]

Projects in the pipeline

Captive solar grew from 300.5 MW to 373.3 MW in the year to June 2026.^[1] KenGen's 100 MW / 100 MWh battery project is at feasibility, with a KES 10.27bn budget, KES 18.55m spent and 2% progress.^{[6] [21]} Industrial projects are mostly small or self-supplied: an 18 MW agreement with Synergetic, a 3 MW proposal from Maxim Agri/Samakgro, and a planned 165.4 MW captive plant for ammonia and fertiliser using KenGen steam. That last project could earn KenGen steam revenue while adding little to Kenya Power's billed sales.^[22] A separate September 2026 framework for about 2 GW of geothermal resource with the same partner is explicitly non-binding.

By FY2035 the three cases give utility sales of 18,712, 24,816 and 16,594 GWh, and peaks of 3,532, 4,444 and 3,500 MW. The distributed case has 11% less grid energy than the moderate case but almost the same peak: daytime solar does not help much at 8pm. The 2015–2035 national plan forecast a 2023 peak of 2,823 MW; the actual was about 2,170 MW, so the forecast overshot by 30%.^[17] Later plans have done better, but forecast error in this sector has usually run optimistic.

The Investor's Conclusion

Which opportunities, if any, offer attractive risk-adjusted returns at current prices?

This chapter asks which investments, if any, offer returns that justify their risks at today's prices.

Securities available to a retail investor

On 30 October 2026, KenGen traded at an average price of KES 9.90 on 1.59m shares and Kenya Power at KES 21.60 on 0.59m.^[8] Those are the only direct Kenyan ordinary-share routes. Kenya Power's 4% and 7% preference shares show prices that imply 15–23% yields, but no shares changed hands that day; we cannot call those yields available. Umeme is cross-listed in Nairobi, but after the end of its Ugandan concession it is a special situation needing its own analysis. A listed infrastructure fund exists, but at 30 September 2026 it held 100% cash and its renewable deals were still in due diligence.

The Treasury bill benchmark

At the 29 October auction the accepted 364-day Treasury bill rate was 9.0288%. After the 15% withholding tax, deducted upfront, the effective net return is about 7.573%; October inflation was 7.3%.^{[9] [10]} Exhibit 14 asks what KES 100,000 invested on 30 October would be worth after three or nine years in each share, against rolling that bill at today's rate.

On moderate demand, the base cases turn KES 100,000 into about KES 93,250 in KenGen and KES 116,380 in Kenya Power after three years, against about KES 124,510 in bills. Over nine years the figures are KES 155,930, KES 222,890 and KES 192,980. Over three years only the bull cases beat the bill; over nine, Kenya Power's base case does too. Every case that beats it depends on resale multiples and on dividends actually being paid.

The scorecard

What would change our view

1. A tariff decision. An EPRA determination for 2026/27 onwards that sets Kenya Power's allowed recovery, including losses.
2. Collections. Two consecutive half-years of falling receivables, especially government and rural balances.
3. Olkaria I terms. Disclosure of the tariff, accepted net output and commercial operation date.
4. Funded capex. KenGen identifying committed finance and contracted revenue for its growth spending.
5. Imports. An amended Ethiopia contract and the voltage-support works that let more than 200 MW flow securely.
6. Open access in practice. The first large customers switching supplier, and the wheeling charges they pay.
7. Reliability. SAIDI falling back towards FY2025 levels while losses keep improving.

Track KenGen, Kenya Power and the rest of your NSE holdings in one place, with alerts when a signpost above changes.

Know someone who would find this useful? Share this report on WhatsApp →

Method, definitions and limits

This report draws on nine research workstreams and an integration review that reconciled 57 conflicts between sources and checked the financial models independently. We report conflicting official figures side by side rather than average them. Financial years run July to June. Money is nominal Kenyan shillings unless stated. Model outputs are labelled "scenario" and use frozen, reviewed inputs; interactive exhibits only switch between pre-computed cases.

• Not available, and not estimated: capacity available at peak; hourly import and curtailment data; the Olkaria I tariff, net output and commercial operation date; KenGen cash-flow statements for FY2015, FY2021 and FY2022; actual October 2026 balances for either company.
• Unresolved conflicts we preserve: Kenya Power's and EPRA's loss ratios (18.58% vs 21.38% for FY2026); utility sales vs category consumption (332 GWh); whether a 30 October buyer receives KenGen's KES 0.75.
• Series breaks: generation mix switches from Kenya Power purchases to EPRA net delivered in FY2025; peak demand from Kenya Power to EPRA in FY2025.
• Model limits: KenGen's base case relies on KES 8.53bn of uncommitted funding and an unbridged KES 17.1bn of growth spend; Kenya Power's value depends on an analyst margin ceiling, not a regulatory entitlement. Two configuration issues in the issuer models affect custom dates and dividend bases only; the published fixed cases are unaffected, which is why custom inputs are disabled here.

This report is general research prepared by Tathmini Research for information only. It is not investment advice, a recommendation, a solicitation or an offer to buy or sell any security, and it does not take account of your objectives, financial situation or needs. A "Watch" disposition describes the state of our evidence, not a recommendation. It relies on public documents listed below, some read through mirrors, and on our own models, whose outputs are conditional on stated assumptions. Figures may contain errors. Share prices can fall as well as rise and dividends are discretionary. Consult a licensed financial adviser before investing. Tathmini Research holds no position in KenGen or Kenya Power and has received no payment from any company discussed.

Megawatt (MW)

A measure of power at a moment: the size of a plant or of demand. 1 MW = 1,000 kW.

Gigawatt-hour (GWh)

A measure of energy over time. 1 GWh = 1 million kWh, the units on your bill.

Nameplate capacity

The rated output of equipment. Not what it can produce at any given hour.

Peak demand

The highest load the grid actually served at one moment in the year. Kenya's comes in the evening.

Curtailment

Output a plant could have produced but was told not to, usually because demand was too low.

Captive capacity

Generation owned by a factory, farm or home for its own use, outside the utility's sales.

IPP

Independent power producer: a private company that builds a plant and sells its output to Kenya Power.

PPA

Power purchase agreement: the long-term contract setting what Kenya Power pays a generator.

Capacity payment

Paid for keeping a plant available, whether or not its power is used.

System losses

Energy bought but never billed: lost as heat in wires, or to theft and metering errors.

SAIDI

Average hours of interruption per customer. EPRA reports a monthly average.

System operator

The control centre that decides second by second which plants run so supply matches demand.

Interconnector

A high-voltage line linking two countries' grids.

Availability-based payment

Paid for keeping an asset ready to use, not per unit carried.

Open access

The right of large customers to buy power from someone other than Kenya Power, paying to use its wires.

Net metering

Crediting rooftop-solar owners for power they send to the grid. In Kenya, at half a unit per unit exported.

Hurdle (discount) rate

The yearly return an investment must beat to be worth doing. Future money is reduced at this rate.

Net present value (NPV)

All future cash in and out of a project, converted to today's money at the hurdle rate. Positive means it beats the hurdle.

IRR

Internal rate of return: the yearly return at which a project's NPV is exactly zero.

Break-even tariff

The price per unit at which NPV is zero.

Profit after tax

Accounting profit for shareholders after all costs and tax. Not the same as cash.

Operating cash flow

Cash generated by running the business, before capital spending and debt.

Expected credit loss

The provision a company sets aside for bills it expects never to collect. Changing the estimate moves profit without moving cash.

VWAP

Volume-weighted average price for the day. Not necessarily a price you could trade at.

P/E ratio

Share price divided by annual profit per share.

DCF

Discounted cash flow: a valuation adding up all future cash available to the company's funders, reduced for time and risk, then subtracting debt.

Treasury bill

A short-term loan to the Kenyan government, bought through the Central Bank of Kenya. The usual low-risk benchmark.

Sources

1. EPRA, Energy & Petroleum Statistics Report for the year ended 30 June 2026 (28 Sep 2026): Table 2.1 p.13; §2.2.1 item 7 p.26 (Olkaria I PPA variation); §§2.1.2–2.1.8 pp.14–19; reliability and losses pp.22–25; Table 2.2 p.17; pp.34–39. epra.go.ke (PDF)
2. EPRA, Energy & Petroleum Statistics Report for the financial year ended 30 June 2025 (29 Sep 2025): pp.13–25, Table 2.2 p.18, Figure 2.5 p.16. epra.go.ke (PDF)
3. Kenya Power, Integrated Annual Report and Financial Statements 2024/25: statements pp.130–135; notes 7–8 pp.167–179 (purchase costs); ECL p.183; Tables 1–1B pp.255–260; customers p.265. kplc.co.ke (PDF)
4. Kenya Power, Annual Report and Financial Statements for the year ended 30 June 2026 (authorised 17 Sep 2026): statements printed pp.55–58; notes pp.80–126; operating data pp.17–25. Read via broker mirror; signed official copy from the issuer's AGM page agrees. mirror (PDF)
5. Kenya Power, "Kenya Power posts KShs 24.99 billion profit after tax…" results announcement (18 Sep 2026). newsroom.kplc.co.ke
6. KenGen, Integrated Annual Report & Financial Statements for the year ended 30 June 2026: projects pp.73–84; auditor pp.120–128; statements pp.129–135; revenue note 6(a) p.163; debt and FX notes pp.185–198; Appendix II pp.246–247; ten-year review pp.249–253. kengen.co.ke (PDF)
7. KenGen, "KenGen Invests Ksh.15.5bn in Power Infrastructure as Kenya's Electricity Demand Rises" (29 Oct 2026): AGM dividend approval, Olkaria I completion, peak demand. kengen.co.ke
8. Nairobi Securities Exchange, Daily Price List, 30 Oct 2026: p.1 Energy & Petroleum; p.2 corporate actions. nse.co.ke (PDF)
9. Central Bank of Kenya, Treasury bill results, issues dated 2 Nov 2026 (signed 29 Oct 2026), p.1. centralbank.go.ke (PDF)
10. KNBS, Consumer Price Indices and Inflation Rates, October 2026 (30 Oct 2026). knbs.or.ke (PDF)
11. KETRACO, Transmission Master Plan 2026–2045: executive summary pp.ii–iii. ketraco.co.ke (PDF)
12. KETRACO, Africa50–PowerGrid transmission PPP project disclosure (project agreement 15 Dec 2025), pp.1–3. ketraco.co.ke (PDF)
13. World Bank, Eastern Electricity Highway Project, Implementation Completion and Results Report ICR00049 (29 Jul 2024), pp.6–11, 18–20. worldbank.org (PDF)
14. Kenya Gazette Vol. CXXV No.73, Gazette Notice 3899, Schedule of Tariffs 2023 (24 Mar 2023), Parts II–III, printed pp.1279–1288. Read via a licensed distributor's copy. copy (PDF)
15. EPRA, public notice cancelling consultative forums after withdrawal of the 5th tariff control period application (published 5 Jun 2026). epra.go.ke (PDF)
16. Ormat Technologies, 2025 Form 10-K (26 Feb 2026): Item 1A and Item 7, Kenya Power receivables. SEC filing
17. Ministry of Energy & Petroleum, Least Cost Power Development Plan 2024–2043 (June 2024): Table 3-11 p.53; Table 5.6 pp.76–77. pppkenya.go.ke (PDF)
18. Energy (Electricity Market, Bulk Supply and Open Access) Regulations, 2026, Legal Notice 79 (8 May 2026), rr.7–12, 18–42. epra.go.ke (PDF)
19. Energy (Net-Metering) Regulations, 2024, Legal Notice 104, rr.6–11. epra.go.ke (PDF)
20. IRENA, Renewable Power Generation Costs in 2025 (July 2026), Figure S1 p.14. irena.org
21. KenGen, Battery Energy Storage System project page. kengen.co.ke
22. Kaishan Group, release on the KenGen-steam-supplied captive power and fertiliser project. kaishancomp.com
23. Kenya Gazette Notice 155, designation of KETRACO as system operator (signed 21 Dec 2021, published 14 Jan 2022). Online copy not linked.
24. Tathmini Research calculations and models, valuation date 30 Oct 2026: capacity boundary arithmetic; household bill reconstruction; Olkaria I screening model (R03); demand scenarios (R05); KenGen and Kenya Power cash-flow models (R06, R07) as repaired and audited in the R10 integration review; investor-return illustrations (R09/R10). These are our analysis, not source documents; inputs are drawn from sources 1–22.
25. John Mutua, "KenGen upgrades Olkaria geothermal plant in bid to ease electricity imports", Business Daily (29 Oct 2026), opening paragraphs. Secondary news report: the KES 15.6bn cost and the 45 to 60.5 MW rating are KenGen statements as reported. The 60.51 MW rating is corroborated by EPRA's FY2026 statistics report, p.26, §2.2.1 item 7 [1]. Article ↗
26. NuPEA (Nuclear Power and Energy Agency), J. Mwangi, "Exploring the Prospects of FNPPs in Kenya", presentation hosted by the IAEA, slide 7: table of generating-plant coordinates. Published site points, not surveyed; five small hydros without a published location are omitted from the map. iaea.org (PDF)