Timescale confusions in solutions for the crisis

Originally appeared on the Daily FT

By Prof. Rohan Samarajiva

A few days after the tsunami, I was called to an expert meeting at Temple Trees by the then Prime Minister, Mahinda Rajapaksa. I was seated next to Arisen Ahubudu, the famous giver of names. He stated that we had lost too much territory, including Madagascar, and that we could not afford to lose more. He proposed building a wall around the country, using the traditional techniques used in protecting tank bunds, the ralepanawa. I was stunned that such a nice and well-meaning person could come out with such arrant nonsense. He had confused geological time with human time. 

Timescale confusions of a smaller magnitude are evident among many proposing solutions to our current multi-faceted crisis. 

Solutions to power cuts

We all experience the problem. Some of us understand the cause: no dollars to pay for fuel for the generators that make up for the shortfall from lower production from the hydro generators and Norochcholai. Even if we had the dollars, such fuel is priced in dollars and subject to price fluctuations that we cannot control. It is common sense that we should shift to electricity produced by renewable sources such as solar and wind. 

The problem is that under current market and technology conditions, both the distribution network (low voltage) and the transmission network (high voltage) are limited in how much solar- and wind-generated electricity they can accept. We can, and should, increase the use of electricity from renewable sources, but we need to upgrade the transmission network to be able to do so. Solar panels yield electricity when the sun is out (not at night and not when clouds pass over the panels); the wind will produce electricity even in evenings when our use is highest, but it is still intermittent. Batteries are not cost-effective yet.

Given the need to balance supply and demand of electricity in real-time caused by lack of cost-effective storage technologies, we need a large and modernised system in order to absorb more energy from these intermittent sources. We need to invest in upgrading the national grid and possibly connect to the large Indian grid. Feasibility studies must be done, and investment mobilised. It will take several years for the desired outcomes to be achieved. Increasing solar- and wind-based energy is not a viable solution for our immediate problems, though it is a solution in the long term. Within the applicable timescale, what we need are dollars for coal and diesel.

Promotion of manufacturing

Twin deficits, exacerbated by recent economic mismanagement, caused the crisis. More exports would have addressed the current-account deficit and may have helped with the fiscal deficit if the right tax policy was in place. Roughly $ 11 billion was earned from the export of goods such as apparel, tea, and value-added rubber products before the pandemic. Around $ 7 billion was claimed from service exports such as tourism, software and business process outsourcing. 

It is true that the East Asian Tigers and China took their people out of poverty through the production of goods for export. One has to ask why Sri Lanka (and to a significant extent, the rest of South Asia) failed to ramp up the production of goods for export, relying more heavily on service exports. One could even argue that the apparel industry is a service industry. A tailor who makes a suit out of material given to him is undoubtably a provider of services. The Sri Lankan apparel industry, which is the largest importer as well as the largest exporter, is doing what a tailor does, at scale. If it is manufacturing, it is manufacturing lite.

Until the market opening in 1978, the answer to the question of why we had no industries was that our private sector was weak and lacked capital. Therefore, the State went into manufacturing: steel, plywood, tyres, sugar, paper, shoes, cooking implements, etc. were all produced by fully State-owned enterprises under protection. They produced shoddy goods at high prices for the local market and lost enormous amounts of money. The plywood factory resulted in the clear-cutting of half of Sinharaja. After the market was opened to imports, they went out of business.

Since 1978, we have relied on private investors, with or without foreign partners, to manufacture for export (and for domestic use). They have tended to invest in sectors that did not rely too heavily on cheap energy (because our electricity prices were high, especially for industrial users). Except in the case of a few sectors such as apparel and rubber-based products, our producers failed to secure access to markets. Restrictive laws and para tariffs hindered local producers from getting integrated into global production networks, with very few exceptions. 

So, the industrialisation prescription as a solution to the crisis will take time and effort to implement. We would have to ensure reliable and low-cost energy (and other infrastructure services such as waste disposal), eliminate para-tariffs, and create the conditions for market access. The latter is the most challenging. 

Investors such as Michelin ensured market access for the solid tyres produced in Sri Lanka. The apparel industry also benefited in the early stages from foreign investors who facilitated market access. Attracting such investors and entering into trade agreements are needed for market access. But both take time. 

Industrialisation may be a good solution, but it is not for the Government to decide on manufacturing priorities. Because China has established itself as the factory to the world, countries such as ours must identify and exploit niches. Those best positioned for this are those with intimate knowledge of the markets, with skin in the game, namely private investors. The State must create the conditions and leave the actual investment decisions to such players. All this will occur on a timescale different from what is relevant to emerging from the present crisis.

Constitutional reforms

It has become evident that the hyper-presidential system created by the 1978 Constitution has failed to yield the promised benefits and has caused serious damage after the enactment of the 20th Amendment, which removed all the checks that were placed on the President by the 19th Amendment. For example, the Minister of Finance has stated that specific officials were responsible for the tax cuts that triggered the present crisis and the delay in debt restructuring. In the current system, the sole authority for those appointments was the President who must therefore be held accountable for the current crisis.

To address the demands of the protestors, the President must go. He must resign or be impeached. The former can take place immediately would allow the country to return to normal (if such a condition exists after the devastation wreaked by the President and his appointees). The time taken to impeach will be too long. 

The next best solution is to reduce the powers of the President. This would require a Constitutional amendment. An amendment that is approved by Cabinet can be completed within around six weeks. If it is moved as a private member’s motion, it could take more than six months, outside the timeframe needed to calm the country and get the debt restructuring done. The announcement that the Government is proposing the restoration of the 19th Amendment suggests a solution within the required timescale. Of course, it would be necessary to scrutinise the proposed amendment and ensure the President’s powers are meaningfully reduced immediately.

In innumerable discussions I have participated in, I hear proposals for Constitutional reform that pay no heed to the time factor. Some talk of a Constitution authored by the people, modelled on what is going on in Chile. The process began with an amendment to the Constitution and a referendum in 2020. This was followed by an election for a Constituent Assembly in April 2021. Its deliberations are ongoing. How realistic is this kind of process for the kinds of issues that have brought our people to the streets?

In these days of limited attention (and paper supplies), it would be useful if greater weight is given to the appropriateness of the proposed solutions for the time needed to solve the problems that beset us.

Rohan Samarajiva is founding Chair of LIRNEasia, an ICT policy and regulation think tank active across emerging Asia and the Pacific. He was CEO from 2004 to 2012. He is also an advisor to the Advocata Institute.

How can affordable electricity be assured 24x7?

Originally appeared on the Daily FT

By Prof. Rohan Samarajiva

The best way to understand the value of something is to experience life without it. These days, the Government is giving us a crash course on the value of reliable electricity supply. An unpleasant lesson, but nonetheless a learning opportunity.

If we probe the causes of load shedding, the learning can be deeper. Load shedding can be eliminated but at a cost. When hydropower declines due to periodic drought, the difference can be made up with generators running on imported fuel, the dollar price of which is determined by world market conditions. We can have 24x7 electricity, but not at an affordable price.

The Government created the immediate conditions for unreliable electricity supply through mismanagement of the country’s external debt. Today’s problems are not caused by delays in building additional generating capacity; they are caused by the lack of dollars to provide fuel for the existing generating plants. But there were deeper weaknesses in the organically developed system that must be understood.

With benchmark crude oil prices going over $ 100 per barrel, we must rethink our dependence on imported fossil fuels.

Reducing dependence on fossil fuels

Examination of the composition of our imports (Figure 1) shows that refined petroleum and crude oil taken together is the largest or second largest category of what is imported. It follows then that reducing the import of petroleum products would be an action that would satisfy many: those concerned about global warming will be made happy; those who want self-sufficiency would also be pleased. 

Petroleum imports are not used solely for electricity generation. But the way to reduce the consumption of petroleum products for transportation also involves electricity generated by renewables: buses and trains that are powered by electricity; lorries, cars, three-wheelers, and two-wheelers that are powered by electricity. Promoting electric vehicles makes no sense unless electricity comes from renewable sources. 

The significant increase in expenditure for fossil fuels starting in 2011 (Figure 2) appears correlated with the massive increase in the vehicle stock after the end of the conflict, leading to a doubling by 2014. Luckily, the biggest increase was in two wheelers, which do not take up a lot of road space and consume less fuel. 

Generating electricity from renewables does require some imported elements such as low-cost, efficient turbines and photo-voltaic panels but the costs and dependence is nowhere near that which exists with imported oil and natural gas. In fact, it may be possible even to export electricity at certain times of the day or even for months on end. But this will require substantial investment in the transmission grid.

Preconditions for increasing use of renewables

An economics commentator whose work I follow had expressed puzzlement at “demand for electricity is higher than supply” being given as a reason for load shedding. Others had expressed outrage at some Facebook posts that I had shared, which stated that solar and wind could not provide a complete solution to our energy woes. These responses by well-meaning and intelligent commentators made me realise the need for a better understanding of how the electricity is generated, transmitted, and distributed.

For all practical purposes using currently affordable technology, electricity must be treated as something that cannot be stored (but see discussion of pumped storage below). That means that it must be generated at the same time as people consume electricity by activating lights or appliances. Peak consumption in Sri Lanka (in the evening hours starting from around 6:30 p.m.) is around 2 or 2.5 times that of lowest use which is around 1000 MW. 

That necessitates a cheap source of baseload electricity that can be drawn upon throughout the day. In addition, we must have other sources that can be mobilised as demand increases. One would think that the major hydroelectric plants that have been built on the main rivers which generate cheap electricity that is unaffected by world market prices and the value of the rupee could serve as the source of baseload power. But there are constraints, such as competing demands from agriculture. The weather affects hydropower, as we are experiencing now. 

Therefore, planners in the past argued for coal as the ideal baseload for Sri Lanka. If Norochcholai does not keep breaking down and operates optimally, it can give 900 MW continuously whether or not the rains come. But it does break down, and it appears there have been irregularities in coal purchases. Coal, even if procured on long-term contracts at the lowest possible price, still must be paid for in dollars.

There are those who argue that Sri Lanka has plenty of wind and sun, and we can solve all problems by shifting to wind and sun. But the simple fact is that these are intermittent sources. Solar does not produce electricity when the sun does not shine and produces less when clouds cover the sun. Wind can produce throughout the day and night, but there are times when the wind dies down. It requires complex system controls to blend these intermittent sources into a centralised system designed for large, stable and controllable generators. 

Countries have incorporated massive amounts of intermittent renewable sources. In 2019, 47% of Denmark’s electricity came from wind. But they have a very sophisticated grid that is capable of handling intermittent power sources, and they use interconnections with other national systems to help balance the system. So, for example, when excess power is generated by the Danish wind turbines, it is used to pump water back up into reservoirs in Norway and Sweden (a method of storing electricity in the form of water known as pumped storage), which can then be run through turbines again to produce more electricity when needed. Yet with all that, Danish consumers pay more for electricity than their neighbours.

Similarly, if Sri Lanka is to increase the use of intermittent power sources, we will have to upgrade the grid and the system control centre’s software. Given the difficulties of synchronising the frequencies to one big plant such as Victoria, it may even be necessary to gradually convert the grid to direct current. If the Sri Lankan grid is connected via a high voltage direct current cable to the Southern Indian grid, the much larger combined system can absorb a greater amount of wind and solar power. 

Interconnecting does not mean that a country gives up on generating its own electricity. It simply means that marginal amounts of electricity will flow in either direction when it is advantageous to two (or more) systems. The fact that the peaks are different in the two systems can also be used to reduce the high costs incurred at peak.

It may be necessary to directly link revenues derived from regulated prices to those who make the substantial investments needed for the grid. This will almost necessarily require a restructuring of the current ungainly, unresponsive, and money-losing CEB in a manner that allows the transmission unit to be run efficiently. 

All these options require careful study in terms of costs, benefits and energy security. The relations between Denmark and its neighbours are such that all the parties can be confident about the contracts being respected and any disputes that arise being settled in a fair manner. We must ensure that the interconnection agreements with India have all these safeguards. The precedent of India’s interconnections with Bhutan shows that mutual interdependence is achievable in South Asia. The experience in Europe where interconnection, including over long distances across water, is growing rapidly even after Brexit, will have to be studied. 

Rohan Samarajiva is founding Chair of LIRNEasia, an ICT policy and regulation think tank active across emerging Asia and the Pacific. He was CEO from 2004 to 2012. He is also an advisor to the Advocata Institute.

Reforming Sri Lanka’s power sector

By Ravi Ratnasabapathy

 The article originally appeared in the Daily News.

Electricity was introduced to Ceylon by a private company in 1895, but since 1927, with the formation of the Department of Government Electrical Undertakings the industry has been a vertically integrated state monopoly.

The electricity infrastructure comprises generation, transmission and distribution. Transmission refers to the bulk transfer of electricity from power plants to substations located near demand centres. Distribution is the delivery of power to consumers from substations.

Some reform of the industry took place during the 1980's and 1990's. LECO, a state owned private company established in 1983 to undertake the distribution of power in Kotte. Independent Power Producers (IPPs) and small hydro developers entered the industry in the mid1990s when generation was opened to private investors following a severe power crisis in 1996.

Since 2004 policy reverted to state-lead investment with the exception of small renewable power projects. The CEB reports regular losses, is heavily indebted and has invested billions but does Sri Lanka have an efficient and economic system of electricity supply, the stated mission of the CEB?

The disaster that is the coal power plant is well known and provides good reason to reassess the long term plans for the provision of power.

Before examining long term solutions there are immediate problems that need to be addressed so some short-term measures are necessary. A peculiarity of Sri Lanka's electricity demand is the high evening peak load. A steep increase in demand occurs between 6pm-7pm which then peaks from 7pm-8pm. Thereafter demand gradually eases over the following three hours.

Peak demand is about 50% higher than average demand and coping with this presents the most urgent problem for the CEB. A study by the Public Utilities Commission of Sri Lanka (PUCSL) in 2012 recommended that “aggressive action is still required to curb further growth in peak demand, since an adverse trend is observed during recent past”.

The simplest solution to this is to move to daylight saving time, which means setting the clock forward by an hour. This proved to be an effective curb on demand when it was implemented after the power crisis of 1996. It was previously used in Ceylon during WWII to conserve power and also by Pakistan after a power crisis in 2008. It is a simple cost free solution that demands immediate implementation.

The management of the demand for power by bulk consumers is also needed. An overlooked aspect of this is the waste of power in the telecommunications industry. Transmission towers consume a lot of power but operators in Sri Lanka do not have a comprehensive infrastructure sharing regime. Operators regard their networks as a source of competitive advantage and share only limited sites. This has resulted in widespread duplication of infrastructure, unnecessary strain on the grid and unsightly visual pollution. The Telecommunications Regulatory Authority (TRC) needs to impose a proper infrastructure sharing regime. Sharing must cover all infrastructure including SLT’s fibre backbone and the TRC should incentivise the decommissioning of redundant sites. Moving on to longer term solutions should private power have a role to play?

In Sri Lanka IPPs have been controversial but the solution is not be to ignore the private sector but instead to move to electricity auctions to procure power. Auctions increase the competition and transparency of electricity procurement and are now quite widely used. Examples include the UK, New Zealand, Australia and Singapore. Open, transparent competition promotes efficiency and reduces costs to consumers. Singapore moved from state monopoly in 1995 to competitive market in stages over a period of years, yielding tangible benefits to consumers. Although the price of oil, the major cost in electricity generation, increased by 152% between 2001 and 2008, Singapore’s electricity tariff rose only 14% during that period. This was possible largely due to efficiency gains in generation, such as utilising more cost-efficient technology.

Competitive pricing encouraged firms to invest, for example in more efficient gas fired combined cycle turbines and retro-fitting existing plants. The share of electricity generated in Singapore by natural gas increased from 19% in 2000 to 79% in 2010 and overall power generation efficiency increased from 38% to 44%. Consequently, carbon dioxide emissions per unit of electricity generated declined by 30% between 2000 and 2007.

The tangible benefits from liberalising the electricity market make a compelling case to move in that direction although the process is by no means a simple or easy. Even in Singapore the major reforms were introduced gradually over a period of a decade, but this should be the vision for Sri Lanka’s power sector.

The Pathfinder Foundation published a paper in 2007 examining in some detail how Sri Lanka could move to a competitive electricity market. The conceptual model for an electricity market, in very simple terms is to have the generation, transmission and distribution split into independent units with competition between them.

It is essential to have several entities carrying out generation (IPPs and entities carved out from existing CEB generation assets). These will compete in a daily computerised auction to sell power to the transmission entity, which should have no links to the generating entities. The auction is usually held a few says before the actual despatch of power is needed. The generating units compete to supply power for fixed time slots in the day, usually for each hour or half hour and the system automatically awards the time slots to each generating unit based on the lowest cost. The transmission entity in turn sells the power to distribution units, which will be monopolies in their respective areas of operation. When distribution utilities operate in similar operational areas, the regulator can easily set up realistic performance targets by comparing their performances. Similarly since the transmission will be carried out by a separate entity the losses in transmission are easily monitored and the incentive is created to minimise leakage.

If practical advice were needed the Government of Singapore has always been willing to share its expertise.

---

Ravi Ratnasabapathy trained as a management accountant and has broad industry experience in finance. He is interested in economic policy and governance issues.