Peace for the World

Peace for the World
First democratic leader of Justice the Godfather of the Sri Lankan Tamil Struggle: Honourable Samuel James Veluppillai Chelvanayakam

Sunday, July 16, 2017

Don’t be fooled by obfuscations Dummies guide to solar, wind and grid power


Daily output variation of a large solar power plant in Spain (The messiness is due to passing cloud cover)

by Kumar David- 

Controversies frequently erupt around renewable energy availability, prices and incorporation in the grid. Much is couched in language opaque to the newspaper reader. That’s a pity because the citizen eventually foots the bill and suffers the power cuts. Since I have some slight familiarity with electricity I am drafting a primer to help layman decode experts and wag a finger at their obfuscations.

The starting point is an appreciation of the difference between power and energy; in our context MW (power) and MWh (energy). One MWh =1000 kWh; kWh is a fancy name for a ‘unit’. If you have a 480 horse-power Ferrari you sure have lots of power but if your petrol tank is near empty you don’t have much energy to go places. A big MW solar plant is not much good if there is frequent cloud cover like a Ferrari with an empty tank. When people con you with solar and wind and talk in MW (power), ask about energy, the MWh expected per year. Lanka has a limited number of onshore wind sites and is nowhere near as prolific in insolation (sunlight falling on the earth) as the Arabian, Thar or Mohave desserts or the bright and cloudless Chilean Andes. We don’t have vast uninhabited lands for photovoltaic or wind farms unlike China and the US; no sites for large, say 500 MW, projects as in Pakistan or India. The 100 MW first stage of Quad-e-Azam solar (final 1000 MW) in Pakistan needed 500 acres! An average wind speed of 5m/sec at 50m height is poor, over 7m/sec is good; few places in Lanka meet the requirement.

The technical catch is that coal, liquefied natural gas (LNG) or nuclear, if well maintained and if fuel supply is not disrupted, can run flat out say 85% of the time (capacity factor). There are 8,760 hours in a year so a 1MW plant can generate 0.85x8760x1 = 7446 MWh average a year. The story with wind and solar is different. The sun does not shine at night and seasonal cloud cover bucks expectations. The wind blows when and where it listeth. Solar and wind plant may provide full output, absolutely no output, and much of the time partial output. The capacity factor for solar and wind at a good site may reach 35%, humbler ones may be constrained to 20%. Therefore, a solar or wind farm will provide only 25% to 40% of the energy of a thermal plant of the same name plate rating. A 1MW solar or wind plant at a location with capacity factor say 28% will provide an average annual energy of 0.28x8760x1 = 2453 MWh. So, talking megawatts (MW) alone is a smokescreen.

The government foolishly dumped the proposed 500 MW coal station at Sampur and recently turned its back on a 1200 MW Japanese LNG proposal in the Western Province. Instead the Energy Ministry has opted for a 50 MW solar plant at Sampur. If we calculate as in the previous paragraph, it has dumped 3.7 TWh of Sampuor energy, and has declined 8.9 TWh of Japanese LNG energy for a mere 0.12 TWh of solar electric energy (1 TWh = 1 million MWh). Incredible! The silly sods in the line ministry, SEMA and PUCSL have set us up for a massive shortage of future electric energy. Pie in the sky talk of 2,600 MW of solar and wind power by 2037, even in the MWs materialise, will not meet a fifth of the country’s then electric energy needs.
The within-day, daily and seasonal energy availability of solar and wind electricity fluctuates. When the sun shines effective costs are good thanks to rapid technological and price advances. The best sites, globally, are competitive with coal and LNG – say Rs 9 to 14 per kWh. The problem is the capital cost of additional (reserve) plant needed when solar/wind power is unavailable. This pushes the effective energy cost up by say 30 to 50% depending on the assumptions made in calculations.

Do the rich subsidise the poor?

There is a peculiarity about our tariffs. Services like state health and education are paid out of national revenue into budget accounts and are evenly priced for all. Similarly, at the retail store you do not pay a different price for sugar, rice or thalana batu depending on your consumption, income or social class. It’s different with CEB electricity. Domestic premises consuming 90 units a month pay Rs 9.25 per unit averaged over all component charges; if consumption is 200 units this rises to about Rs 23 per unit and at 400 units a month it is a shade over Rs 33 per unit. So, there is a commodity specific cross-subsidy; bad in principle and bad in practice. People get the municipality to provide different assessment numbers for upstairs and downstairs and pay Rs 9.25 a unit on each bill instead of Rs 20+ or 30+ rupees a unit on a composite bill. This tariff system should be dumped. The supply authority should divide revenue needed by units expected to be sold and charge everyone at a uniform rate. It is mockery to attempt social justice on a commodity by commodity basis.

The silliness does not stop there. The tariff on industrial and commercial customers is complicated depending on time of day (this is reasonable since generation cost at peak time is high), consumption and type of industry. It varies from more than Rs 25 at peak time to less than Rs 10 from late night to early morning. But it is irrational to have so many classes of consumers.

The craziest is religious premises. The charge is less than Rs 2 per unit for the first 30 units and even at 200 units (average middle-class household consumption) it is less than Rs 10 a unit! There is a big cross-subsidy. Subsidising less well-off households is one thing, but why the devil should I pay to have the gullible misled with myths and superstitions? Let the dayakayas and parishioners pay.


When the sun shines bright on your old Kadawatta home the electric energy produced by a roof-top array may be more than needed at that moment. The rating of plant of course is matched to the highest insolation at midday. The thing to do is to store up the extra energy for later. The wind and solar power injected into the Lankan grid in the foreseeable future will be so small that the system can absorb it without flinching (without storage). A few hundred MW when the sun is blazing and wind howling is easy to absorb in a four thousand of MW system. No special measures such as smart-grids or pumped storage are needed. It is different in Holland, Denmark and Germany where the sudden cessation or surge of huge amounts of wind power can panic system controllers. Wind all over a country will only rarely shut down or surge up all at the same time but to deal with big fluctuations the system needs fast acting plant (gas turbines for example) that can kick in quickly. These spinning and standby reserves have to be entered on the cost side of a wind or solar balance sheet. Storage systems for renewables, especially large grid connected plant are now receiving a lot of attention.

The term net-metering is not used for grid level stuff but for consumer premises. That solar panel on your roof may be yanking out full power at midday when no one is home and only the fridge is running. In the evening, your wife turns on the rice cooker, kids switch on study lights and you relax in a brightly lit room with a well earned scotch after a hard days work! (OK, call me an MCP!). Now you need lots of electricity. The trick is a meter which keeps count of the energy injected into the CEB and sets it off against consumption when demand is high – hence the name net. You are billed for net consumption and a surplus can be carried over as a credit (‘banked’) for months or years.

This is where the dispute about cross-subsidies arises. I said a big consumer may pay an average of say Rs 30 a unit. If he sees the light (poor pun) and goes solar maybe his net consumption can come down to 90 units, Hey presto, he pays at Rs 9.25 a unit! So, the argument goes that the CEB loses lots of money and its ability to cross-subsidise poor consumers is undercut. Is this correct? Well yes, revenue is top-sliced and something will have to be raised to recoup it. The most expensive plant is run at peak time but these rich solar types do not top-slice their demand at these times. Hence the cost saving to the system from their ‘returned’ energy is small.

This objection is invalid if time of day pricing (a more sophisticated version is called ‘spot-pricing’) is used. Credit for energy returned and charge for energy consumed depend on the utilities cost of generation at the relevant moment. If you inject 100 units at noon when, for example, the utilities’ marginal cost is Rs 12 a unit, your credit is Rs 1200, but if 100 units in consumed during the evening peak when the utility is running its plant flat out at a marginal cost of Rs 25 per unit, the debit side of your account will show an entry of Rs 2500. You are in energy balance but not in money balance.

Since our solar and wind penetration will remain minuscule for a long time it is unlikely to make much impact on overall generation costs. It is good in principle to encourage renewable energy but not to rush in like a bull in heat, create future energy shortages and fall prey to the rapacious oil lobby – which is where we are heading. Trump is cuckoo to spurn the environment; our half-baked experts and perplexed regulators are dyslexic and foreswear a measured approach. A US Energy Administration graph showing how coal usage gradually declined is reproduced. Other forms of fossil fuel usage have not declined, but that’s a story I should take up another time.