Philippine Electricity Markets

This Chart is a reminder that we’ve seen higher monthly...

nicknich3 — Wed, 20 Jun 2012 15:46:00 +0800

This Chart is a reminder that we’ve seen higher monthly average WESM prices before.

I’ve plotted the Monthly Average WESM price here on the blue line (placing it at the 15th of each month) and superimposed some daily average prices in the background.

But I’m focused on the monthly average, which for May was P6.8/kWh. That’s high for 2011/12 but in 2010 we saw monthly prices average P10.5/kWh and even P13/kWh.

When looking at monthly averages, it’s not a game of trends so much as it is a game of possibilities. You can look at the chart and see that for the next few months we have a very wide range of possibilities. We know from history that prices exceeding P10 on a sustained monthly basis are perfectly feasible.

2010 was a time of much different coal prices than we have now. Much different. But high WESM prices are not driven by coal prices. - it’s driven by oil utilization. Which in turn is a result of a tightening in the demand/supply situation (which is hardly price related at all).

By the way, we’re more than half way through June and prices so far this month are averaging about P8.3/kWh.

These are updated through June 10. On the top chart, I’ve...

nicknich3 — Wed, 20 Jun 2012 07:26:00 +0800

These are updated through June 10.

On the top chart, I’ve added the 200 day moving average so you can tell when daily prices move above or below the long-term average. You can also see the daily average price spike on June 8/9 when we had the weekend problems with Sual and Ilijan and prices spiked to the P55-60/kWh range for a few hours.

The bottom left chart shows that this time last year prices were already headed down from their typical May highs - but not so far this year.

The bottom right chart shows that demand, as typical this time of year, is starting to head back down to its trend line after the May highs. We’ll have to see how that goes.

Photo

nicknich3 — Sun, 17 Jun 2012 19:17:48 +0800

BP Statistical Review of World Energy 2012

nicknich3 — Thu, 14 Jun 2012 15:05:14 +0800

BP Statistical Review of World Energy 2012:

Released today. Addressing the year 2011. Lots of data available.

“Global energy consumption grew by 2.5% in 2011, broadly in line with the historical average but well below the 5.1% seen in 2010. Emerging economies accounted for all of the net growth, with OECD demand falling for the third time in the last four years, led by a sharp decline in Japan.”

“The averages hide a mixed picture by fuel, however. Oil demand grew by less than 1% - the slowest rate amongst fossil fuels - while gas grew by 2.2%, and coal was the only fossil fuel with above average annual consumption growth at 5.4% globally, and 8.4% in the emerging economies.”

How Much Energy is Meralco Taking From the Spot? It has steadily...

nicknich3 — Thu, 14 Jun 2012 08:06:00 +0800

How Much Energy is Meralco Taking From the Spot?

It has steadily trended down to about 5% now.

Monthly Power Supply Cost Updates for May The May WESM...

nicknich3 — Wed, 13 Jun 2012 07:44:00 +0800

Monthly Power Supply Cost Updates for May

The May WESM Settlements Price is my estimate - the actual hasn’t been released by WESM yet.

Saturday Update on Generation Status The table shows Friday and...

nicknich3 — Sun, 10 Jun 2012 08:32:00 +0800

Saturday Update on Generation Status

The table shows Friday and Saturday back-to-back. Hourly delivery is shown as a percent of unit’s rated output.

You can see we lost Sual 1 right after midnight Friday. It came back on-line at 4pm Saturday and began ramping back up to full load over the next four hours.

You can see Ilijan cut back to 50% (aggregate) output 10pm Friday then down to 30% at 2am Saturday, then back to 45% by 8pm Saturday.

Sual 2 still out continuously.

We lost Calaca 2 at 4pm Saturday but it re-started at midnight.

Sta. Rita 6 (which is actually San Lorenzo 2) began having problems at midnight Friday and we essentially lost it completely at 5am Saturday.

I can’t detect any problem with gas supply limitations at Sta. Rita/San Lorenzo - so if there is indeed a gas supply issue, it seems to be limited to Ilijan.

What Caused the Shortage of Supply Last Night? Here’s a...

nicknich3 — Sat, 09 Jun 2012 16:01:00 +0800

What Caused the Shortage of Supply Last Night?

Here’s a news article with quotes from Meralco.

The table above is mine where - I massaged data downloaded from the WESM web site. It shows the hourly generation at the Luzon coal and gas plants, by hour, for yesterday.

Note that as of 9pm, all of these major units were running at essentially 100% of their capability - except for Sual 2 which has been down for the past three days and Calaca 1 not shown because it’s been off-line since October of 2011 for rehab. Calaca 2 was slightly de-rated.

Something happened after 9pm, because Ilijan, which had been running at it’s full rating of 1,200 MW dropped back to 600 MW, or 50% of its rating.

Now, WESM hasn’t published the file for Saturday yet. So I can’t tell what happened after midnight. But apparently at 12:32am Sual 1 tripped off-line. That’s a huge loss. The total Luzon load [note I’ve corrected this] at that time was around 5,700 MW. So that’s 11% of the Luzon load.

Also, the article cites “limitations of supply” at Sta. Rita and San Lorenzo, which at midnight were running full capacity. So whatever happened at Ilijan (such as a gas supply constraint) may have spilled over to the First Gas plants after midnight? Note that one of the units at San Lorenzo started feeling the pinch at midnight - it dropped by 85 MW from the previous hours.

I note that we’re also setting historical high load levels over the past 60 days - both on-peak and off-peak.

Hi Nick! You're doing so well here! I was wondering if there is (will be) a 'transmission loss factor' or any Philippine Standard (an oxymoron!!!)? I am aware there is a 'systems loss cap' closely monitored by ERC which accounts for non-technical and technical losses. Thanks, Troy

Fri, 08 Jun 2012 15:30:54 +0800

Hi Troy. The Systems Loss Cap you mentioned refers to losses on the Distribution system. I read your question as asking me if there is something similar for the Transmission system which is owned and operated by NGCP. Correct?

Frankly, I havent’ dug deeply into ERC’s regulation of NGCP. They are under performance-based ratemaking similar to the private DUs. I also know that they retain very capable outside, international consulting expertise to assist in the regulation of NGCP. I don’t know how transmission systems loss enters into the regulatory equation.

I also note that actual systems loss on the transmission system is not only a function of the physical transmission network, but also a function of generation dispatch directed by the Market Operator (PEMC) and DU loads.

Bottom line: I don’t know the answer.

"This June, you will notice a 71-centavo/kWh increase in your electric bill. While Meralco did not..."

nicknich3 — Fri, 08 Jun 2012 14:23:00 +0800

“This June, you will notice a 71-centavo/kWh increase in your electric bill. While Meralco did not increase the rates for distribution charges, there was a rate increase in the Wholesale Electricity Spot Market (WESM) due to the unavailability of two of the biggest and cheapest energy sources: the Sual Power Plant run by San Miguel Energy, and the Pagbilao Power Plant run by Aboitiz Power.”

- Quote from Meralco

Here’s a look at what the two Sual and the two Pagbilao units were doing in the May billing month.

Sual 2 was down about four days, Sual 1 struggled for 3 or 4 days, and Pagbilao 1 was offline for a week. Here is a chart showing the daily average MW from these plants and I’ve overlaid the daily average WESM price.

Interestingly, these four units generated more kWhs in May than they did the previous month. And there’s not a lot of correlation between the capacity line and the WESM price bars in the chart above. The issue is not quite as simple as just Sual and Pagbilao unavailability, but it is a significant component.

Here is a chart of May WESM prices in the context of the two prior months. I’ve plotted the average monthly price too.

Meralco’s WESM costs depends on when they were in the market and for how much volume. Presuming they were primarily in the market during Peak period hours, you can see that peak prices took a whopping 91% jump from P4.00/kWh in April to P7.65/kWh in May. But that’s partly because April was a low month compared to March.

Another issue is that I took a preliminary peek at WESM load levels and if I’m not mistaken, loads may have increased by as much as 10% in May compared to April. That in and of itself will tend to move WESM prices up quite significantly, regardless of Sual and Pagbilao.

So there are a number of things going on. I haven’t totally digested it yet.

Coal Prices keep falling. The June 6 closing prices on Newcastle...

nicknich3 — Thu, 07 Jun 2012 13:53:00 +0800

2012 Forward Curve w/ Actuals

Percent Annual Change in Forwards

Coal Prices keep falling.

The June 6 closing prices on Newcastle Futures point to an average 2012 price of $99/MT by my estimates. That would be an 18.4% drop from 2011 actuals.

And we’re looking at 2013 prices of only a dollar more at $100/MT.

Deconstructing the WESM Price Setters

nicknich3 — Sun, 03 Jun 2012 15:53:00 +0800

One of my favorite things to look at is who is setting the marginal price in each hour. I’ve updated my Marginal Plant analysis to incorporate a full five months of data for 2012.

I’ve added two columns to my plant listings [see tables at end of this post]. In addition to the Frequency that each plant is selected as the Marginal Plant and the Average MCP it bids when selected, I’ve added columns to show the Maximum MCP it bid when selected and the Coefficient of Variation in its MCP when selected.

As a review, Coefficient of Variation is the ratio of the Standard Deviation in MCP to the Average MCP and is a measure of how much variation around the mean (expressed as a percent of the mean) we see in MCPs.

The first table is sorted by Frequency.

The second table by Avg. MCP.

The third table by Max MCP.

The fourth table by Coef. of Var. in MCP.

The last two tables show the trend in Frequency and Avg. MCP by month for each plant.

Looking at Prices

Focus on the second table for a moment; the one sorted by MCPs. All of the top plants are oil-based peakers, obviously. You have to go all the way down to position #11 to find a non-oil unit. And guess what it is? San Roque, a hydro plant, at P9.4/kWh.

The next three non-oil plants are also hydro, before you get to the first coal plant which is (could you guess?) PEDC on Panay at position #20 at P4.7/kWh. PEDC is priced a full P1.8/kWh above the next highest coal unit, CEDC on Cebu, and P2.5 above our cheapest coal unit Masinloc.

MCP Variability

I expected the the coefficient of variation (cv) to be reveiling. It is to some extent. The top plants with high cv tend to be peakers or hydro - those entities that are trying to capture market share at the upper end of the price curve. Hydro is tricky because only a portion of their capacity is playing that game. So hydro can have high cv and low mean price. But in general, I submit that high cv indicates the entity is trying to price opportunistically. PEDC is a case in point. Interesting that CEDC is also high on the list. But Dingle is puzzling (PDPP and PDPP3). It’s winning bids within a very tight price range.

Unified Leyte is leading the list with the highest cv. Intersting that none of the Green Corp geothermal plants are ever priced at the margin. Thus they don’t appear on my lists.

I need to do more analysis on the cv aspect.

Here are the tables.

Dual-Settlement Concept Model for Utility Bilateral Contracts

nicknich3 — Sat, 02 Jun 2012 17:17:00 +0800

Before coding the calculations required to model a bilateral contract between a distribution utility (DU) and an IPP, you have to have to have a conceptual model of the transaction in mind. And there are multiple different ways to conceptualize the settlements of a bilateral contract in the Philippine markets.

I find that the literal calculation methods PEMC uses in their actual settlement procedures to not be very useful and prefer something I’m calling a “duel-settlement” concept model.

Both the IPP and the DU have, conceptually, two separate settlements.

A real-time settlement (with WESM)
A BCQ or bilateral settlement (with each other)

My conceptual model is mathmatically equivalent to the PEMC settlement methodology. But the PEMC methodology conflates what is essentially a financial transaction with a real-time product purchase of actual electricity. I like to separate the two so we can see what’s going on from a financial value standpoint - so it helps planners. It has the added benefit of revealing and clarifying negotiable options and flexibility in the contracts - helping negotiators on both sides.

#1: IPP Real-Time Conceptual Settlement

The IPP settles it’s real-time transactions with WESM based on whether it was scheduled by WESM as a result of its hour-ahead WESM bids.

If WESM schedules the IPP, the IPP receives the market price (at its node) for that hour and incurs its own dispatch cost to generate (fuel + variable O&M). It’s net position that hour (excluding fixed costs which are essentially sunk costs) is the difference in those two items and is typically a positive value. I call this the “operating margin”. For gas-fired plants it’s referred to as the “spark spread.” For coal plants it’s sometimes referred to as the “brown spread” or “dark spread” for oil. It’s simply the margin by which market prices exceed price of fuel (and variable O&M) but is never less than zero since the plant would never bid a price lower than it’s variable cost to operate (including shut-down/start-up costs).

If WESM doesn’t schedule the IPP, the IPP receives zero revenue from WESM and incurs zero cost. It’s net position is always zero in such instances.

Note that over a year (or whatever time period), a merchant IPP wishes to have its “net operating margin” be enough to cover its fixed costs (including capital recovery profit margin on investment) and do it in a manner that doesn’t create unmanageable cash flow problems.

#2: DU Real-Time Conceptual Settlement

The DU real-time settlement is simple: it purchases its total hourly load requirement from WESM at whatever price exists in the WESM for that hour.

#3: IPP Bilateral Conceptual Settlement

The IPP receives the hourly equivalent of the bilateral contract fixed payment, whether the DU nominates MW for that hour or not. If the DU does nominate, then the IPP also receives a variable or dispatch payment from the DU. And if the DU does indeed nominate, then the IPP also has to purchase those MW from WESM to deliver to the DU under the bilateral (it’s a fictitious delivery, not a real one).

#4: DU Bilateral Conceptual Settlement

The DU pays the hourly equivalent of the bilateral contract fixed payment to the IPP, whether the DU nominates MW for that hour or not. If the DU does nominate, then it also pays a variable or dispatch payment to the IPP.

Illustrative Model

I have an Excel spreadsheet which you can download from here that provides two small tables summarizing these settlements under various market price conditions and assumptions on bilateral contract price terms. If the link doesn’t work for you, let me know.

Let’s look at some charts generated by that model.

The first chart above shows the Total DU Cost and Total IPP Revenue across a range of market prices. Note that the Total IPP Revenue from its two settlements (real-time and bilateral) is a constant P2.5/kWh. The IPP is totally indifferent to WESM market prices. That’s what the bilateral contract (hedge) does for it. It guarantees a P2.5/kW/hour revenue - which has presummably been designed to cover its capital recovery and profit objectives.

The first chart also shows the DU Total Cost which ramps up from P2.5/kWh to P5.0/kWh and is then capped at P5.0, regardless of WESM market prices. Again, this is the objective of the DU to limit it’s exposure to spiking WESM prices.

Uncovered Positions

Let’s look at what would have happened to the DU and the IPP had they not entered into this bilateral. The following chart shows the net total revenue to the IPP with and without the contract cover. As you can see, when prices are low, it receives more revenue under the contract than it would have without the contract and when prices are high, vice versa. At a market price of P5.0/kWh, the IPP is at break-even.

The following chart shows outcomes for the DU with and without the bilateral contract. When prices are low, the DU pays more under the contract than it would without it and when prices are high, vice versa. At a market price of P5.0, the DU also is at break-even

Mark-to-Market Positions

For each of the prior two charts, we can look at the difference between the solid and dotted lines. That tells us, for each market price point, how much money the IPP or DU lost or saved with the contract as compared to without the contract (mark-to-market position).

For example, for prices below P2.5, the IPP always makes P2.5 more than it would have without the contract where it would have made zero (since the plant is idle at those prices). Similarly, for prices below P2.5, the DU always pays P2.5 more than it would have without the contract - reflecting the payment of fixed costs in those hours when it did not nominate and utilized lower cost WESM power instead. The contract also sat idle for them.

As you can see, the DUs mark-to-market position is the exact mirror image of the IPPs.

In a sense, the DU is making a bet that prices will predominately exceed P5.0/kWh and the IPP is making a bet that pricess will predominately fall below P5.0/kWh.

Well, not exactly. Because even if prices didn’t predominately exceed P5.0, the DU may be happy to have taken the bet so that it never was exposed to prices exceeding P5.0. And even if prices predominately exceed P5.0, the IPP may be happy to have taken the bet so that its revenue never went below P2.5.

Breakout! Well here’s the expected breakout on WESM...

nicknich3 — Mon, 28 May 2012 08:53:59 +0800

Breakout!

Well here’s the expected breakout on WESM prices, confirming the uptrend I discussed in a previous post. WESM hasn’t published the monthly averages yet for May, but they have published the daily averages through May 20.

But even before any May prices were published, the technical setup - the triangle as I’ve drawn - guaranteed a breakout one way or the other. And we all knew it pretty much had to be to the upside, given the season.

But it’s a pretty big breakout. The question is will it end up exceeding the October recent high of P6.56, which I would view as meaningful. I didn’t draw the connector line - you can imagine it. Given the nature of these types of technical setups, I wouldn’t expect a return to sub-4 peso monthly prices for quite some time. We’ll see.

The Coming Crash in WESM Prices First off - I don’t...

nicknich3 — Sat, 26 May 2012 14:14:13 +0800

The Coming Crash in WESM Prices

First off - I don’t believe it’s likely to happen this year. But it seems inevitable over the next two years or so. As I noted in my previous post, load is growing at almost 6% per year. The price trend line in the chart above indicates that prices are growing at a 30% per year clip. That results in a price elasticity of demand calculation of 23% - which isn’t sustainable, probably, in our economy.

Two possibilities: We’ll see demand destruction or (more likely) we’ll get the expected new entry of several new base load plants. The entry of that new supply (whose total capacity is expected to exceed new load growth over the next 2-3 years) is expected to drive down spot prices and arrest the medium-term average rate of increase measure until demand growth again squeezes reserve levels and we get another cycle of up and then down prices.

It’s the nature of electricity spot markets.

But that doesn’t mean retail electricity customers are exposed to this whip-saw effect in prices. Their prices are determined not by the spot market per se but rather by the bilateral supply contracts their supplier has entered into. These contracts, many of which are long term, smooth out the end-user price from year-to-year and season-to-season.

Which also means that end-use customers are not going to feel the full effect of the inevitable price collapse in wholesale spot prices whenever it does come - and quite possible very little of it.

Luz/Vis Load Growth The slope on the Average Demand trend line...

nicknich3 — Sat, 26 May 2012 11:07:00 +0800

Luz/Vis Load Growth

The slope on the Average Demand trend line (data is for the past seventeen months or so) indicates that load has been growing at an average rate of 5.9% per year.

I also note the typical “Easter Drop” and “May Recovery” as I’ve circled for last year and this year. This year, both the dip and the recovery was a bit more pronounced than it was in 2011. An indicator of more robust industrial activity?

You can also see the Christmas drop from this year. I didn’t have enough data to capture that for the previous year.

Hi Nick - Do you know of any restrictions concerning the supply and installation of VAWT Systems? Thanks, Ed Skinner.

Sun, 20 May 2012 15:18:46 +0800

Hi Ed. Well, first there may be a site-specific restriction on wind. Each proposed installation must first conduct a Grid Impact Study (GIS) that is reviewed by NGCP to assure that the installation will conform with the Philippine Grid Code and/or to identify capital improvements to the grid or special operating procedures that will bring it into compliance. The GIS looks at both steady state and transient stability issues.

Second, as part of the FIT Hearing process, the ERC is reviewing NGCP’s feasibility study of renewables integration into the grid nationally and will be guided by that in specifying an initial cap of the amount of Wind, in general, that will be be eligible for the FIT. But that ERC restriction, whatever it is, would not restrict Wind in general, just the amount that is eligible in the first phase of FIT. VAWT could always negotiate for a bilateral arrangement with an off-taker, subject to GIS restrictions.

Hypothetical Mindanao Price Premium Mindanao utilities are short...

nicknich3 — Sun, 20 May 2012 15:03:00 +0800

Hypothetical Mindanao Price Premium

Mindanao utilities are short of power during certain peak hours. What if Mindanao utilities could buy power from the grid at the same prices that Visayas utilities can purchase during peak hours? How much would they be able to save in that hypothetical situation compared with the Therma Marine barge price offers?

It depends, of course, on how often they need the peaking power.

Looking at Q1 2012, the highest hourly price that occurred on the Visayas WESM grid was P47/kWh. And there were only five hours in which prices exceeded P30/kWh. Thereafter, prices dropped rapidly. The red line on the chart shows the average price one would have paid for power had they taken power at prices equivalent to the top 10% of the hours, 20% of the hours, etc. all the way up to 100% of the hours.

For example, if you had taken power at the grid prices for 100% of the time, you wold have paid an average of P3.8/kWh. If you had taken power only during the top 20% of the hours, you would have paid an average of P9.0/kWh.

The blue line is indicative of how Aboitiz has structured their barge pricing for early 2012. If you utilized your barge contract 20% of the time, for example, you would pay about P13.5/kWh for that peaking power.

Using the 20% time as an indicative benchmark for how utilities are utilizing the barge contracts (it varies by utility), that P13.5/kWh price represents, then, about a P4.5/kWh premium over the hypothetical situation where they could have purchased at Visayas WESM prices of P9.0/kWh.

You can see that the barge pricing structure very closely approximates the shape of Visayas grid prices. However, it exceeded the Visayas grid prices across all utilization factors, with the premium being highest at the lowest utilizations. The premium is the least somewhere around 30% utilization and rises slightly as utilization rises from that point forward.

NPC's New Rate Increases - In a Nutshell

nicknich3 — Fri, 11 May 2012 11:34:00 +0800

The Decisions are here and here.

The ERC has determined that over the past 4 years, NPC has under collected certain costs related to generation charges and exchange rate movements (GRAM and ICERA).

So it turns out that ratepayers over those years (2007-2010) have received a subsidy and ERC will permit NPC to recover those losses from future ratepayers who will have to pay a premium.

Total amounts subject to recovery from future ratepayers are:

Luzon: P35.1 billion

Visayas: P8.5 billion

Mindanao: P1.7 billion

The collection will be like this (my estimates):

Luzon: P3.7 billion/yr over next 8 years and P2.6 billon/yr over subsequent 2 years

Visayas: P0.9 billion/yr over next 5 years and P0.7 billion/yr over subsequent 5.5 years

Mindanao: P0.4 billion/yr over next 4.5 years

Recovery started in April. Here’s what NPC rate trends have looked like in P/kWh. You can see the April uptick in this chart:

Keeping a sense of perspective: A few selected Apr-2012 price...

nicknich3 — Fri, 11 May 2012 07:33:08 +0800

Keeping a sense of perspective: A few selected Apr-2012 price levels compared with Jan-2008 (about four years ago).