Revisiting Texas's Electrical Power Predicament—Part 2

I ended Part 1 with questions about the future of older coal fired electricity power plants and the possibility of replacing plants and expanding electricity demand as Texas's economy and population continue growing.

The demise of coal-fired plants

Back in late 2011, the owner's of Monticello coal-fired plant in Titus County, Luminant, had threaten to shut down or "idle" rather than implement costly upgrades  to put the plant in compliance with the Clean Air Act, and, ERCOT suggesting that they couldn't force Luminant to keep a plant open in violation of federal pollution rules (see e,g,m Texas power grid operator says blackouts possible). 

What happened to these and similar coal fired plants? 

It appears that the Monticello plant was mothballed for the winter of 2012.  The plant capable of producing 1,880 MW, enough to power 0.94 million homes in normal conditions, apparently was not needed in winter months and therefore was not given a “Reliability Must Run” designation by ERCOT.  After re-opening for the summer of 2013, the plant again was mothballed for the winter, and Luminant requested to also mothball its Martin Lake coal-fired plant for the winter (A Changing Market and Dim Future for Coal in Texas).  Martin-Lake has a capacity of 2,250 MW enough to power 1.12 million homes.  Altogether, Luminant owns five coal-fired plants with a name-plate capacity of 8,017 MW—I wonder if eventually all five of these plants will get shut down during non-summer months. 

Although part of Luminant's reason for mothballing the plants is stiff economic competition from power generated by natural gas plants, I think that it is the cost to comply with the EPA rules under the Clean Air Act is what will lead to the eventual demise of all of these plants. 

Indeed, last summer, the DOJ, at the behest of the EPA, filed a law suit against Luminant alleging continued violations under the Clean Air Act, citing the Martin-Lake plant and the Big Brown Power Plant, located near the Dallas-Fort Worth area (Luminant hit with suit alleging Clean Air Act violations).  Luminant has argued for special protection against the Clean Air Act for violations during startups and shutdowns,and  the Fifth circuit agreed, but then, the denial of an appeal to the Supreme Court doesn't bode well for eliminating the possibility of subsequent law suits attacking this theory. 

Additionally, CPS Energy, in San Antonio, appears to on track with its plan to shut down entirely the JT Deely coal-fired plant at the end of 2018, citing as the reason the high cost for environmental retrofits to comply with new emissions requirements. The Deely plant which has a power generating capacity of 871 MW will be replaced with a natural gas-fired plant. 

A new coal fired plant, the Sandy Creek plant near Waco, with a 900 MW capacity, did come online in 2013.  Its opening, however, was only after a settlement with the Sierra Club who had fired a lawsuit against the plant's owner alleging violations of the Clean Act, among other things.  After years of legal battles, as part of the settlement the owners agreed to pay for stricter air pollution controls and to not build coal-fired plants in Georgia and Arkansas.  The Sandy Creek owners are still embroiled in legal battles with a company, NAES, hired to maintain and operate the plant, alleging that NAES's mismanagement damaged the plant's boiler (Sandy Creek power plant owners suing over 2011 boiler incident). 

It is apparent to me that no additional, Clean Air Act Compliant coal fired-plants will be opened in Texas any time soon (if ever)  due to litigation costs, federal regulations and the present low price of natural gas (After White Stallion Power Plant Canceled, Coal Faces Dark Future in Texas). 

A power vacuum

While some see this as "big win for clean air in Texas," and it probably is, I wonder how this will impact Texas's ability to meet its growing power demand. Are there any non-coal power generating projects underway to actually expand the power grid and not just replace the existing capacity produced by the coal plants being mothballed or shutdown? 

In an interesting Forbes article, Will Summer Blackouts Doom The Texas Boom?, Christopher Helman gave a summary of Texas's energy status going forward.  New electrical capacity from Nuclear Power does not appear to be in the cards given the blockage of to two new Japanese Toshiba reactors.  Two 760 MW capacity gas-fired plants owned by Panda Power are scheduled to come on line in 2014, and there is a third plant scheduled to come on line in 2015.  The 540 MW natural gas-fired Ferguson power plant in Llano in 2014 will replace the old 420-MW plant which closed in the fall of 2013.  Such gas fired plant are important to ERCOT for their ability ramp up quickly when electricity demand spikes—something that wind or solar power, or, even nuclear or coal are NOT particularly good at.   Finally, Helman's article talks about another 3,000 MW of wind power set to be built by 2015.  But even if all of this wind power is built, for reasons already discussed in past articles in this blog, on a hot Texas summer day, one might only expect to get ~10% or less of the name plate capacity from wind.

That seems to be about it—about 1,500 MW in 2014, and another 1,000 to 1,500 MW in 2015, depending on how you feel about the wind power contribution you could count on in the summer.

Keeping in mind that at a peak summer power usage of around 67000- 68000 MW, and, the need to grow the capacity by about 2 percent per year, ERCOT's goal should be to add about 1350 MW capacity per year every year, just to keep it's reserve margin about constant.

But ERCOT's own Capacity-Demand-Reverse Report for 2013 shows their expectation of a shrinking reserve capacity, going forwards (Figure 4). 

As illustrated (red circle), ERCOT expects its summer reserve margin to steadily decline from a margin of about 13.6% in 2014 of total capacity to 4.5% by 2023.  For comparison, the North American Reliability Corporation’s acceptable reserve margin is 15 percent, and, NARC estimate of Texas's reserve margin to be 12.9%. NARC indicated that the peak demand growth rate in Texas (about 2.7%/year) is projected to be the highest in the United States.

It is noteworthy that the growth in year-to-year summer peak demand shown in the above table (blue circle) appears to be based on (or is equivalent to) an assumption of demand growth increasing by 3 percent per year (e.g., 72071 x 100%/69807=103 %) for the next few years, and then, suddenly slowing down to less than 1 percent per year by 2021.   I don't know where these projections come from. 

If I take ERCOT's 2014 "firm load forecast," and increase it by a constant 2 percent per year every year, then the reserve margin drops below 0% by 2023.  If I take that same 2014 firm load forecast and increase it by a constant 3 percent per year every year, then the reserve margin drops below 0% by 2020.  Of course, weather is the wild card, and on any one summer day, particularly hot weather could shift ERCOT into emergency load shedding.  We already know from the experience of 2011, and ERCOT's own emergency plans, that coming within ~3.4 percent of power resource capacity would trigger industrial load shedding with rolling blacks outs to follow soon afterwards.  Texas has just been lucky so far.

ERCOT has done two other things in an effort to mitigate Texas's inadequate and declining reserve power margin.

Understanding that Texas needs the summer capacity from the aging coal-fired plants like Monticello, Martin-Lake and Big Brown, ERCOT has raised its price cap on wholesale electricity, presumable to motivate owners, like Luminant, to de-mothball these plants every summer and supply an important part of the based load summer power capacity.  A coalition of these power providers are pushing for some kind of guaranteed compensation in the form of permanent cap increases to keep these plants available for summer use in the years ahead (Rolling blackouts are Texas' future without reform, generators say)

I don't see this as a workable solution, given that in 2014 the EPA wants to implement more stringent rules for CO2 emissions from existing plants, and, none of the existing plants in Texas are in compliance with such limits.  The owners of these plants, like Luminant, find the costs of retrofitting these plants with carbon capture technology "unworkable." (Texas electric grid getting greener even before EPA crackdown). 

If, for example, Luminant decided that it wasn't financially worthwhile anymore to "de-mothball" the Monticello and Martin-Lake  plants for the summer and just leave them closed for the summer of 2014 and beyond, then that would drop ERCOTs power generating capacity by about 4000 MW. The 2014 reserve margin would then drop to 7.8% percent and to only 4.8% by 2016.  I'm not sure who the public would blame for the likely ensuing rolling blackouts—the EPA, Luminant or ERCOT?  Would the Clean Air Act rules be waived after a few summers with extended periods of rolling blackouts and people dying from heat exhaustion?  Perhaps we will see.

The other thing that ERCOT is doing is promoting voluntary conservation.  In 2013 ERCOT introduced a voluntary pilot program, the 30-Minute Emergency Response Service ("30-Minute ERS), for homeowners groups and commercial users to be compensated for cutting their electricity use during times of peak demand when power supplies are tight and prices spike, if, they can reduce power use by at least 0.1 MW within 30 minutes (ERCOT plan rewards electricity cuts during peak demand).  Presumably this option would be implemented at Energy Emergency Level 1, when the grids' reserve margin drops below 2300 MW.   This pilot program is still being tested, but from a November 2013 report with about 1600 potential participants it looks like the 30-Minute ERS could produce reductions of 100 to 200 MW.    

With over 6 million residential smart meters installed across Texas the potential impact of some form of residential power load shedding program is much larger than this.  Residential load shedding would work best if those volunteering for the program simple have their power load shed automatically as needed by the electric utility via a smart meter. For instance, PUC of Ohio is advertising how smart meters will give customers the opportunity to "assist" PUC through voluntary “load shedding” where PUC will send signals to thermostats and other appliances to adjust the devices’ activity until another signal is delivered to restore normal activity. I would think that a special device would have to be installed so that the power company could selectively remotely control these appliances.   But perhaps a smart meter could be used to more grossly limit the total amount of power flowing into individual households.  A smart grid controlling smart meters has been proposed as a solution to mitigate Pakistan's terrible grid problems by Dell.  Pakistan is currently facing a power generation shortfall of 6000 MW which results in 8-12 hours blackouts throughout the country.

Of course, the "carrot" to volunteer for such programs will be a price reduction, or rebate, when and if your power load gets shed (ERCOT will pay homeowners to conserve electricity this summer).   Low income customer will be attracted to this, and, a partial shutdown of certain appliances is certainly preferable to full-scale rotating blackouts.  Those who can afford to pay a higher price for electricity need not worry about such inconveniences—for now.   

Winter power woes

One final note that involves those those coal-fired power plants that are mothballed over the winter and the reliability of the grid.  I expect that mothballing will be a growing trend if natural-gas prices stay low and rules under the Clean Air Act become increasingly stringent and enforced. 

But mothballing a significant portion of the States' steady base load capacity can start to affect power reliabilty during the winter months.  Last November 2013, a cold winter snap caused Texas to have its highest electricity use on record for the month of November of 46,931 MW on November 26. The previous month, October, had also set a new demand record of 54,710 MW for that month.  The absolute magnitude of this amount is paltry compared to the summer month's daily peak usages of 66,000-68,000 MW, but it is during the fall and winter months that power plants either get mothballed or shutdown for servicing.  Still, despite these records set in October and November, I saw no reports of the Texas power grid being stressed. 

A few weeks later on December 6, about 250,000 people lost power in the Dallas Ft, Worth Area during that week—but that was due to downed frozen overhead power lines from a massive storm of freezing rain. Once again there were no reports of the power grid being stressed. 

On Monday, January 6, another cold snap through Texas did cause ERCOT to issue a Energy Emergency Alert and implement “demand response" mitigation with entities that contract to reduce their electric use when needed.  What was the demand level that cause this alert?  Only 55,486 MW—just slightly higher than that October record.  

Power consumption of 55,486 MW was enough to get to an Energy Emergency Alert Level 2, meaning that the reserve capacity was less than 1750 MW.  Texas was probably less than 1000 MW away from rolling blackouts, I suspect.   At the heart of the emergency was the unexpected loss of 3700 MW of power production capacity from two plants that had equipment failures due to the cold weather.  That caused ERCOT to import about 1000 MW of power from the Eastern USA and Mexico and for wholesale electricity prices to hit the cap limit set by ERCOT. 

While the loss of 3700 MW from these two unnamed plants might have been a "surprise," what was not a surprise was the about 10,000 MW worth scheduled plant shut down in capacity for maintenance and mothballing. 

What about the winter wind?  Well that Monday, the wind capacity at ERCOT's disposal was 17 percent of the 10,400 MW nameplate capacity, providing about 1782 MW or 3.2 percent of the grid's total power generation.  (Role of Texas wind power debated after winter emergency; Did Wind Really Save Texas from Rolling Outages?).  I think this just goes to show, once again, that wind can provide some base load capacity but that base load contribution will variable.  There still has to be enough steady base load (coal and nuclear) and variable (gas) capacity to provide a reliable power margin.  In my opinion, as the coal-fired plants get mothballed in increasing amounts, the size and reliability of this margin will diminish.

This Saturday, January 18, ERCOT issued another Level 1 energy emergency alert after yet another expected/unexplained power plant outage of 1200 MW.  As noted by one reporter, Saturday morning was rather unremarkable weather-wise, with temperatures in the 50s°F.  It is hard to imagine that this plant outage was due to equipment failure from cold weather.  Once again, a Level 1 alert is issued when the margin drops below 2300 MW, so the loss of 1200 MW from one plant was enough to drop the grid below this margin.  If this plant had gone out the week before, then there highly likely would have been rotating blackouts in Texas. 

These alerts underscore just how close Texans are living near the margin for rolling blackouts—trying to walk a tight rope between high prices, federal air pollution regulations and a reliable electricity grid.

===========================

This has been my first "free" weekend since October, where I have not either been sick, working, or, both!  I will be back again, to what has become this occasional blog, as time and health permits.  

4-6-2014: clean up some types and add to legend of figure 4.  

Revisiting Texas's Electrical Power Predicament—Part 1

I have written about Texas's electric power situation a few times over the past few years, and now, it's time for an update. 

I wrote about Rolling Blackouts in the South Western USA due to the shut down of multiple electrical power generating plants in Texas during a cold snap in February 2011, with cascading effects on the ability of neighboring state New Mexico's ability to receive pipeline shipments of natural gas and have rolling blackout of there own. 

As I discussed in August 2011 Revisiting Rolling Blackouts in Texas, rolling blackouts nearly reoccurred, as the State faced what turned out to be the hottest summer on record, for Texas and some neighboring states. The body responsible for managing electric power flow for ~85% Texas, Electric Reliability Council of Texas (ERCOT), estimated that the State came within 50 megawatts (MW) of interrupting flows to industrial customers, that is, industrial load shedding. 

In January 2012 I took a more in depth look at Texas's Electrical Power Predicament with a three part series (see Part 1, Part2, Part3).  Back then, my assessment was that there were three factors all conspiring to squeeze Texas's ability to sufficiently maintain or grow it's electric power supply to avoid rolling blackouts:

            1) increasing demand due to growth in population and increasing economic activity;

            2)  the implementation of several rules by the EPA that would cause some of Texas's older coal-fired power plants to shut-down; and

            3) the unlikelihood of alternative electrical energy sources being built in time or have sufficient capacity to meet the expected summer power demand.

Maybe some think that should have added a fourth factor: ever-increasing hotter summers, but, I did not see this as necessary condition to cause an increasing mismatch in electricity supply and demand sufficient to cause rolling blackouts.   All that would be needed are summers as hot as the summer of 2011, a continuation of the population and economic growth trends, the announced closure of certain coal-fired power plants and a lack in adequate growth or capacity in new power sources.  For instance, in some of the scenarios I ran in my earlier series (Scenarios 1 and 2, Part2), assuming exactly the same energy demand as in 2011, predicted the possibility of anywhere from 12 to 17 of industrial load shedding and from 1 to 11 days of general rolling blackouts for the month of August, that could be applicable to 2012 and 2013. 

Happily, these predictions were wrong.  Most of the rest of this article explores the three above-mentioned factors to assess what did not happen to cause the prediction to fail.

Texas's population, economy and forward predictions

One possibility is that Texas's population growth or economic growth have stopped or at least slowed down since 2011 and therefore the expected demand for electric power was not forthcoming.

The data shown in Figure 1 suggest that this has not been the case.

Figure 1 shows Texas’s population and annual GDP (left and right axis respectively) from 1987-2012 and some projections through 2015.  The blue line and symbols corresponds to the Texas GDP based on US census bureau estimates with projections for 2012 and on from USA Government Spending.   The red line and symbols shows Texas's population from the Texas State Library based on US census bureau estimates.  The green line and symbols shows population estimates and projections from the Texas Department of State Health Services).  The vertical line divides the 2012 and earlier data estimates from the projections for 2013 to 2015. 

From 2011 to 2012, Texas's population according to the US census bureau data increased by 1.5 percent/year.   The Texas DSHS estimate is higher at 2 percent/year.   A 2 percent/year population growth rate is closer to Texas's average yearly rate of 1.97±0.63 percent/year (last 20 year average and SD) using the US census bureau data.  Texas's population growth rate is about double the USA's yearly rate of 1.05±0.58 percent/year (last 20 year average and SD).  The Texas DSHS projections for 2013-2015 suggest continued growth rates of 2 percent/year. 

From 2011 to 2012, Texas's GDP was estimated to increase 5.8 percent percent/year.  That is slightly lower than the 20 year average of 6.3±3.3.  The GDP is projected to increase by 4.4 to 5.8 percent/year for 2013-2015. 

Overall then, there are few signs of any significant downturn or slowdown in Texas's population, economic growth, or, in government expectations of growth going forwards.  In a previous post, I showed that Texas's electric generation capacity has been growing at about 2 percent percent/year to match the population growth rate (Part 1 Figure 3).  I would expect that the growth rate is power generation would have to stay at 2 percent/year just to keep up with the population growth rate. 

Texas's summer weather

Another possibility is that there have been cooler summers in Texas and therefore less electricity demand, since in the summer, air-conditioning can account for a substantial amount of electric power use. 

Data reported by the Southern Regional Climate Center suggests that the summers 2012 and 2013 were not particularly hot, at least compared to the summer of 2011. 

Figure 2 shows the average monthly temperatures for Texas for the last 20 years for the months of June, July August and September (caution the vertical scales for each month are not the same). 

The red circles highlight the average temperatures for 2011, which was high for June, July and August, but not so high for September, as compare to the previous years since 1993 or compared to 2012 and 2013.  In contrast 2012 and 2013 average temperatures are closer to trend line for this period. 

So, electric power consumption in Texas benefited from having typical summer temperatures in 2012 and 2013. 

Texas's summer electrical power consumption. 

Of course, Figure 1 and 2 doesn't tell us about daily high temperatures, which typically occur in the mid-to-late afternoon, and would typical be the cause of electrical power demand spikes as people return to their residence and turn on the AC to cool the house down.

To get a better view of this, you have to look at the actual hour-by-hour power loads. 

I still have the hourly load date for the July August and September 2011, but sadly I did not save the data for June 2011, which is no longer available at ERCOT's website.  I am saving the data for June-Sept 2012 and June-Sept 2013 before these disappear as well.

Figure 3 shows the hourly loads for August 2011 (red) August 2012 (green) and August 2013 (blue).  The solid pink line shows ERCOTs estimated capacity to produce electricity (73000 MW), and, the orange line shows the point where industrial load shedding would begin (68344 MW), circa August 2011.

In August 2011, there were several days in early and mid-August that got close (e.g., within 1500 MW or 66844 MW) of industry load shedding in the late afternoon): Aug 1, Aug 2, Aug 3, Aug 9, Aug 18, Aug 19.  The all-time peak for 2011 was 68293 MW at 4 pm Aug 3.  That was about 50 MW away from load shedding.

In contrast, for Aug 2012, only one day reach similar levels: Aug 1 at 66489 MW.  In fact, this was the only day in the summer of 2012 that exceeded 66,000 MW.

For Aug 2013, there were a few more days where power consumption exceeded 66,000 MW: Aug 5, Aug 6 and Aug 7, with Aug 7 being the highest at 67180 MW—that's 1164 MW less than the all time peak day of Aug 3, 2011.

Summary

Even though Texas's population and economy has continued growing at the same pace as in 2011, its power consumption was not as high in 2012 and 2013 as in 2011.  Probably average summer weather, at least cooler than 2011, had a lot to do with this.  For the last two years, Texas has been lucky by having just average summer temperatures. 

One thing that is apparent to me is that Texas is in a race to grow its power generation at least at the same rate as its rate of population increase, with the severity of summer weather thrown in as a wild card.  

So what about those older coal fired power plants that in 2011 were expected to be shut down under the EPA enforced clean air act and various anti-pollution rules?

In part 2, I will discuss Texas's coal-fired plants and other potentially sources of electricity that may or may not be coming in the near future.
==================
January 19, 2014: clean up the legend to Figure 1 and correct some spelling and grammer issues.