Gencos getting fried

The EMA (Electricity Market Authority) has distanced itself from the Hyflux debacle. It takes the position that the failed Tuas Spring power plant is a commercial decision which has nothing to do with them. That is only partially correct. Of course, a private investment decision is not the concern nor responsibility of EMA. In a deregulated industry, investors enter the market based on price signals and other metrics. Nevertheless, it is EMA that provides the final green light.  An application for a new plant would be reviewed on the basis of a host of criteria, such as investor credibility, technology, whether it fits into the overall plan of the national electricity system, etc. 

In the electricity industry, over and under capacity is undesirable. Given that power plant construction is investment heavy and a very long process, it is the reason why one of the most important functions of a regulatory authority in the industry is capacity planning. Over capacity has been there for a long time, but in 2013 Keppel and Pacific Light added on 1,200 MW of CCGT capacity, and Tuas Spring another 395 MW in 2015. That exacerbated the system and brought it to crisis proportion. With generation capacity at 13,600 MW in 2018 and output of 52,000 Gwh, there is roughly 50% over capacity in the system.

Over capacity means price war. Low power generation prices are bringing gencos to the brink of financial disaster. While consumers may enjoy lower prices from competition, made fiercely lower by over capacity, the state of the industry is untenable. The electricity industry is the underbelly of the economy. Disruption, instability, or collapse of the power generation sector is unthinkable. 

What we have here is yet another government agency that has been inflicted with the disease of accountability-phobia. EMA has a lot of explanation to do on the failure of capacity planning. Whilst the underbelly of the economy is getting fried, parliament is treated to  Oxley 38 soap operas and histrionics of 'Ah Gong and his grandson' that bordered on the juvenile and imbecile. No one has thought it fit to grill the Minister of MTI.

Should you be concerned? Certainly. Power generation needs to be sustainable in the long run. It has to, and will, be resolved. Any solution to prevent a collapse of the industry will most invariably fall on the shoulders of consumers by way of higher tariffs. 


SRMC Vs LRMC  (Short Run Marginal Cost Vs Long Run Marginal Cost):

A quick economics 101 for those without this background. SRMC vs LRMC is something that often confuses even economics students. Most books explain this with geometric models that can go pretty deep. I'll try to explain this in a more intuitive way,

First of all, MC (marginal cost) is the cost of producing one more unit of the product.  Initially MC decreases as more and more units are produced, due to economies of scale.  Up to a point, the Law of Diminishing Utility sets in, and MC rises. A manufacturer will not produce anything beyond the point where the MC is more than the MR (marginal revenue, or sales price) because it will mean a net cash outflow for each additional widget produced.

Under SRMC, the plants are already in operation. All fixed cost have been expended. The only costs to produce the next unit are variable costs -- direct labour, fuel, direct overheads. As long as the price for the product is higher than the MC, production can continue as the product will have net revenue to help recover setup costs or fixed overheads. In other words, SRMC focus is on the cash flow. 

For LRMC, it is forward looking. The plants have not been set-up. So production cost for the project takes into consideration all costs including the plant setup costs. This is the financial information the investors need to decide if the project is viable.


USEP (Uniform Spore Energy Price) vs Vesting Price :

Brief note on vesting contracts : It is a hedging instrument, a contract for difference. Certain quantities of gencos' output are tied up in vesting contracts with SP Services. The parties settle the difference between vesting prices and actual power generation prices.
For detailed explanation of vesting contracts, refer to the previous blog here

The USEP is based on the clearing price of each 30 minute auction in the SWEM (Spore Wholesale Electricity Market). As the gencos offer price is based on SRMC, we can assume the USEP as representative of the market SRMC.

The Vesting Price for the LNG Vesting Contracts and the Balance Vesting Contracts are computed on the basis of LRMC of a hyphotetical efficient CCGT plant. So we can use the vesting prices as the market LRMC.

The chart above shows a very clear picture. Before 2013, the USEP and Vesting Price evens out more or less, as it should be in a real efficiently functioning market. From 2013 up to current time, USEP has been significantly lower than the Vesting Price. This means the gencos SRMC are running below the LRMC. Gencos are not earning enough to cover their fixed cost, and of course, no ROI.


Spark spread:

A cool way to assess genco profitability is by a metric called the spark spread. Energy investment analysts compute this to gauge the financial viability of power generators. A negative spark spread means the genco is definitely loosing big time. The spark spread is basically a theoretical gross margin. It looks at a genco's electricity price and deducts the fuel cost from it. The residue is the spark spread to cover all other non-fuel cost. It's a ball-park figure but rather useful.  The actual computation is difficult as internal data are required. It involves info like fuel hedge ratio, cost of fuel, output load (peak hours distribution), plant efficiency or heat rate, and others. 

Since genco data is not available, we can look at SP computed vesting prices to get an idea of the status. This chart is an estimated spark spread of the LRMC of vesting contracts computed by energy consultant Martin J. van der Lugt. To simplify the explanation, just focus on the blue and orange lines. The blue line is the spark spread. Before 2013 it was sometimes negative and sometimes positive over the orange line for Non-fuel cost. After 2013 it has been consistently running negative. Gencos are running at a loss - the gross margin is insufficient to pay for all other non-fuel costs. (Vesting prices are computed a quarter in advance based on projections. The black line is the spark spread as computed from vesting contract data. The blue line is to sync the projected fuel cost data to realised costs).

Gencos' productions can be split into two parts. One parted is vested, that is, committed to a vesting contract with Singapore Power at a pre-determined price. The above spark spread chart looks at this vested portion and it's showing losses for gencos.   

Balance of the production that is not vested is in worse situation since the electricity generation price, the USEP, is lower than the LRMC. 


Vesting contracts offer gencos some protection:

Vesting contracts as a hedge, may or may not be beneficial to gencos, depending on how the generation prices will be. Because vesting prices are higher than USEP since 2013, gencos gain in the price difference of the strike price of the vesting contracts and the USEP. The higher the proportion of their production that has been committed under the vesting contracts, the better for the gencos as they can look to the hedge for some compensation. The interesting question is -- who has the higher production vested? Well, in order to push the LNG storage project, EMA implemented the LNG Vesting Scheme which binds the new CCGT plants coming onstream in 2013.  Under this scheme, the relevant gencos are bound to purchase certain quantities of their fuel from the LNG terminal. So surprise, Hyflux's Tuas Spring plants, together with Pacificlight and a new plant of Tuas Power, they have LNG vesting contracts. So they actually have some advantage in the market.

LNG Vesting Scheme is here to stay until 2023. So if the authorities want to tweak the system to assist Hyflux in its predicament, this is one area where they can do it. Simply tweak the LRMC model used to set higher vesting prices. If LIBOR can be tweaked, this is a piece of cake. I'm not saying they will do it, but technically this is one way out. If this is done, who is paying the bill? Well the regulated tariff is computed on the basis of the vesting prices plus the adjustment from contract differences.  So non-contestable consumers foot the bills. This is pure speculation, folks. 


Gencos are really bleeding:

The over-capacity is an albatross round the necks of the gencos for years.  Almost all gencos are in the red for the past few years. The industry as a whole is loosing close to a S$1B a year now. It is a crisis situation.










Hey, the SWEM actually works:

The Singapore Wholesale Electricity Market is designed as a free market system where prices will be determined by demand and supply forces. Gencos make their offers every 30 minutes in the SWEM and the supply is auctioned off based on prices. The lowest priced will be accepted first and it goes up the price list until the demand load is satisfied. The higher priced offers are rejected. With over capacity, gencos price themselves low.

The way the SWEM is set up, it will never work in an under capacity environment. The system works well with some over capacity. But with serious over capacity, which is where Singapore is right now, it works extremely well. In an over supply market, gencos are forced to offer low resulting in low USEP. As long as their offered price is still above their marginal cost, gencos will bid in the wholesale auction to try to get the dispatch to keep the plants going. They cover marginal costs, but do not earn enough to cover fixed and other admin costs. Gencos with less efficient plants are forced off the market first. Plants using steam turbine technologies with higher SRMC will drop off the chart. This has already happened.  The bump up in capacity in 2013 caused USEP to drop significantly. 


What is the solution:

Actually, the solution is absolutely simple. Price is a function of demand and supply. It's of course impossible to ramp up massive demand overnight. But supply can be dropped immediately.  Simply retire one or more of the existing plants. The question is, who wants to do national service.

Export the excess capacity, so said some. That means generate the full capacity and export the excess. Unfortunately, the idea of exporting electricity is a piped dream for the moment. The Asean Grid is being discussed at ministerial levels but it will be decades before it can be realised. Cross border sale of electricity is not new but requires lots of protocols, infrastructure, legislation, and treaties. 

In Asean, corruption is high in many countries. Power generation is one great opportunity for greased money. Why should politicians forgo this opportunity to develop themselves and resort to buying from Singapore. Even, or unless, Singapore electricity is extremely low.  Though luck, within Asean, Singapore and Philippines have the highest tariff, due in part both countries carry no subsidies. The other countries have tariffs that are significantly subsidised. Difficult to explain to their citizens why they import high from Singapore and sell low in their country.  

There are lots of political consideration for exporting electricity. Example Germany has excess production which they export via the European Grid. Germany heavily incentivised green energy with feed-in tariff, so consumers pay high electricity at home and the government exports cheap to other countries. So let's say Singapore has the opportunity and exports electricity. What it will be doing, under present circumstances, is to export and reduce supply in Singapore, thereby enabling gencos to push up the USEP. A moral anomaly to export in order to increase electricity prices for local consumers. 

One way or another, EMA will have to resolve the over capacity mess.  It is absolutely unacceptable for those entrusted with the system to say it's not their fault.