One energy system, too many silos

Why Pakistan must stop managing fuels and start governing energy as a system.

Pakistan’s energy crisis is not the result of one failed policy. It is the cumulative outcome of fragmented decisions that made sense individually but created costly feedback loops across electricity, gas, LNG, solar and industrial demand.

Pakistan’s energy sector has spent decades trying to solve one crisis at a time. Yet each fix has too often shifted pressure elsewhere, creating the conditions for the next crisis.

We added generation because there was not enough electricity. We kept gas cheap to protect consumers and industry. We imported LNG when domestic production declined. We promoted captive generation when the grid was unreliable. We encouraged solar when grid electricity became too expensive. And whenever circular debt became unbearable, we raised tariffs.

Almost every decision had a rationale. Many made sense when viewed on their own. Together, however, they have produced a system that is expensive, fragile and increasingly difficult to manage.

The problem is not simply bad policymaking. It is also the way Pakistan thinks about energy. Too often, policymakers maximise one desirable outcome while ignoring how the wider system will respond. But energy systems do not stand still. Change one part and consumers, investors, technologies and institutions adjust. The eventual outcome can be very different from what was intended.

The lesson is simple: the optimum is rarely the maximum.

We built power plants, not a power system

Pakistan once genuinely lacked generating capacity, so building power plants was a defensible response. But generation is only one link in the chain. Electricity must also be transmitted, dispatched, distributed, billed and collected before it becomes useful energy for households and industry.

We expanded generation faster than the networks and institutions required to use it efficiently. The result is an absurdity: paid-for generating capacity that cannot always reach consumers because the rest of the system is constrained.

This is not an argument against building generation when it was needed. It is an argument against treating generation as though it were the electricity system.

The same mistake persists whenever planning focuses on megawatts added rather than electricity delivered reliably and competitively. A power plant without adequate transmission is not an energy solution. It is an expensive asset waiting for the rest of the system to catch up.

Cheap gas ultimately became expensive gas

Gas tells the same story.

For decades, Pakistan treated cheap domestic gas almost as an entitlement. Low prices helped consumers and industry in the short term and avoided politically difficult adjustments. But price is not only what a consumer pays; it is also the signal that shapes consumption, production and investment.

Suppress the economic value of a resource for long enough and two things happen simultaneously: consumption is encouraged while investment in new supply becomes less attractive.

Domestic production then declines relative to demand. The shortage has to be filled through imported LNG, exposing the country to international prices and foreign-exchange risk. A policy intended to preserve cheap gas ultimately helps create dependence on expensive imported gas.

Then the problem compounds. Raise gas and RLNG prices sharply and consumers respond again. Industry becomes more efficient, switches fuels, installs alternatives or reduces production. Throughput falls, fixed costs are recovered over fewer units, and tariffs come under further pressure.

This is the part policymakers repeatedly underestimate: consumers are not fixed entries on a spreadsheet. They react.

Solar is not the problem

The solar boom is the latest demonstration.

For a household or business facing expensive grid electricity, installing solar can be entirely rational. Falling equipment costs and high tariffs have made the economics increasingly compelling.

But millions of rational individual decisions can change the economics of the grid itself.

Electricity networks carry large fixed costs. When higher-paying consumers buy fewer units from the grid, those costs do not disappear; they are spread across a smaller sales base. Tariffs rise, solar becomes even more attractive, and grid sales fall further.

The wrong conclusion is that solar must therefore be stopped or penalised. Solar is not the problem. Consumers are responding to incentives created by the system.

The real question is how distributed solar fits into a modern power market alongside transmission, storage, flexible generation, network charges and competitive supply.

Trying to maximise solar would be foolish. Trying to suppress it would be equally foolish. The objective should be to integrate it.

Circular debt is the scorecard, not the disease

Pakistan also treats circular debt as though it were primarily an accounting problem.

When the numbers become intolerable, the familiar remedies appear: raise tariffs, adjust subsidies, restructure liabilities, inject liquidity and improve recoveries. Some measures are necessary. None, by itself, cures the disease.

Circular debt is the financial scorecard of failures elsewhere: distribution losses, weak recoveries, poor governance, distorted subsidies, rigid contractual obligations, transmission constraints, inefficient companies and, increasingly, demand destruction.

Tariff increases illustrate the trap. Arithmetically, a higher tariff can close a revenue gap. Economically, it changes behaviour.

Industry cuts consumption, loses competitiveness or seeks alternatives. Households conserve or install solar. Grid sales decline while many fixed costs remain. Those costs are then spread over fewer units, requiring still higher tariffs.

At some point the cure starts feeding the disease.

Pakistan cannot tariff-adjust its way out of structural dysfunction.

Efficiency is not the same as resilience

Complex systems need what can look like inefficiency.

A power system needs reserves. Transmission networks need spare capacity. A gas system benefits from diversified sources. Storage may sit idle until it is needed. Competitive markets require multiple buyers and sellers. A country may sensibly retain options that are not the cheapest option every day.

These are not necessarily inefficiencies. They are insurance.

A system designed only for maximum efficiency under normal conditions can become dangerously brittle when circumstances change. Pakistan’s energy policy, however, is frequently designed around the crisis of the moment.

Short of power? Build generation. Short of gas? Import LNG. Electricity too expensive? Encourage solar. Grid demand falling? Protect grid sales. LNG demand falling? Find buyers for LNG. Circular debt rising? Raise tariffs.

The immediate problem changes. The policymaking instinct does not.

AI could make us better — or simply faster at being wrong

Artificial intelligence adds a new dimension to this problem.

AI can be extraordinarily useful for Pakistan’s energy sector. It can improve demand forecasting, dispatch, renewable integration, predictive maintenance, theft detection, grid balancing and investment planning. It can model interactions at a scale no human planning team can match.

We should use it.

But AI is also an extraordinarily powerful optimisation machine. It will optimise the objective we give it, not necessarily the outcome society actually needs.

Ask it to minimise generation cost and it may undervalue resilience. Ask it to maximise renewable penetration and it may produce a technically impressive answer that fails unless transmission, storage and flexibility are properly represented. Ask it to maximise collections and it may recommend measures that improve a utility’s accounts while damaging affordability or productive demand.

The danger is not only that AI might give the wrong answer.

It may give a brilliant answer to the wrong question.

For Pakistan, where energy planning has long been fragmented between fuels, institutions and ministries, that is a serious warning. Better algorithms cannot compensate for badly defined objectives.

Stop optimising silos

Pakistan does not have separate electricity, gas, LNG, petroleum, renewable-energy and industrial-energy crises. It has one interconnected energy system.

Our administrative structure encourages us to forget that.

Raise electricity tariffs and gas demand may shift. Raise gas prices and electricity demand changes. Raise both and solar becomes more attractive. Solar then alters grid demand. Lower grid demand increases the cost per unit of maintaining the network. Higher energy costs weaken industrial competitiveness. Lower industrial production changes energy demand again.

Everything connects.

Yet policy is still too often developed ministry by ministry, fuel by fuel and crisis by crisis.

Pakistan needs whole-system energy planning. Electricity, gas, LNG, coal, LPG, petroleum, transmission, storage, distributed generation, electric vehicles and industrial demand should be modelled together.

That requires institutional reform. If Pakistan wants to overcome fragmented decision-making, it should move toward one integrated energy ministry rather than separate power and petroleum silos. The same logic applies to regulation: electricity, gas, petroleum and emerging energy markets need regulators that coordinate around the system as a whole, rather than agencies optimising narrow mandates.

Leadership matters as much as structure. Pakistan’s energy institutions too often lack professionals with broad experience across power, gas, petroleum, markets, finance and industrial demand.

At the same time, bureaucrats and politicians have treated parts of the energy system as separate fiefdoms to control rather than as one interconnected system to optimise. That culture has strangled innovation and made a system’s approach harder to pursue. A sector that needs experimentation, coordination and market development is still governed too often through permissions, files and administrative discretion.

Coal and LPG should be understood through the same lens. Coal is not merely a question of cheap generation, nor is LPG merely a stopgap when piped gas is unavailable.

Each affects logistics, foreign exchange, household behaviour, industrial competitiveness, emissions, grid demand and investment incentives. Yet Pakistan still relies too heavily on administratively set prices and permissions instead of developing real energy markets. Without competitive gas markets, open access, wheeling, transparent network charges and price signals that reflect scarcity and cost, policy remains an exercise in control rather than coordination. The result is predictable: distorted prices, weak investment, fuel switching by accident rather than design, and another round of crisis management.

And every major policy should be tested for second- and third-order consequences.

What happens to grid economics if distributed solar doubles? What happens to LNG demand if industry electrifies faster than expected? What happens to coal dispatch if imported-fuel prices move or carbon constraints tighten? What happens to LPG demand if piped gas becomes scarcer or more expensive? What happens to power demand if electric vehicles become mainstream? What happens to gas exploration if producers gain greater market access? What happens to tariffs as the best-paying consumers reduce their dependence on the grid?

Those are not side questions. They are the policy.

The optimum is not the maximum

Pakistan’s energy debate is full of demands for more: more generation, more renewables, more gas, more investment, more efficiency, more recovery and more competition.

Most are desirable. None is a strategy on its own.

The objective is not to maximise any single resource, institution or metric. It is to build an energy system that delivers reliable, affordable and secure energy, keeps Pakistani industry competitive and enables the transition to cleaner technologies.

That requires trade-offs.

Efficiency must coexist with resilience. Affordability with financial sustainability. Renewables with flexibility. Markets with effective regulation. Domestic resources with diversified imports. Investment incentives with consumer protection. And increasingly, machine optimisation with human judgment.

For decades, Pakistan has tried to repair its energy sector one component at a time. Too often, individually rational decisions have combined to produce collectively irrational outcomes.

AI may finally give us the tools to understand this complexity. Or it may simply allow us to repeat the same mistakes faster and with greater precision.

The technology is new. The mistake is not.

Pakistan does not need to maximize each energy resource in isolation. It needs to optimise the energy system as a whole.

Copyright Business Recorder, 2026