Seeking Alpha

Glenn Doty's  Instablog

Glenn Doty
Send Message
Doty WindFuels is a subgroup of Doty Scientific - a small company founded by my father 30 years ago. We are currently developing a new energy paradigm - a process for using variable renewable energy to convert CO2 into liquid hydrocarbon fuels and chemicals. The products would be called... More
My company:
Doty WindFuels
  • Recent Analysis From The Union Of Concerned Scientist Is Garbage.

    The Union of Concerned Scientists has recently published a report entitled "State of Charge", which is getting a great deal of press. While I consider myself a fan of much of the Union's work, and I myself have been a financial contributor to the Union of Concerned Scientists in the past... their work is not above reproach, and on occasion in their passion to advocate what they clearly believe there are oversights in their work which are significant enough to lead to completely false conclusions.

    "State of Charge" is, unfortunately, a poorly considered study that uses incorrect methodology which leads to an incorrect conclusion.

    The fundamental problem with the report stems from the basic assertion that if a grid had 5% wind, 1% solar, 20% hydro, 30% nuclear, 20% natural gas, and 24% coal... then that exact same "grid mix" could be credited to the EV which is being plugged in and directly compared to an ICE vehicle which is not plugged in.

    There is a logical fallacy here. When you plug in a NEW demand in to the grid, there must be additional energy provided to the grid. That is not a matter that is up to debate. Just to be clear - I'm stating that when you plug in your vehicle to the grid, demanding another ~20 kWh/night or so, then the power companies must provide an additional 20 kWh/night that they would not have had to provide had you not plugged in your car (had your car run on gasoline, for instance).

    Power companies tend to use the energy source with the lowest marginal cost that can fulfill the need. This again is obvious and is not something that is open to debate. The MARGINAL cost for solar, geothermal, hydropower, and wind is essentially $0, and the marginal cost for nuclear power is ~$5/MWh. That is - the cost of the energy once the capital has been spent.

    To illustrate: Once a dam is in place, the difference in cost between the dam producing energy and the dam just allowing the water to flow past the turbines is literally pennies/MWh. So the power company is, ALWAYS, going to use every bit of hydropower that it possibly can whether you plug your car in or not.

    That means that if you DON'T plug your car in, 100% of the installed hydropower is used. When you DO plug your car in, then the power company must satisfy that additional ~20 kWh of demand by ramping up something that does not include wind, solar, geothermal, nuclear, or hydropower.

    Now we get to the idea of SPARE CAPACITY. The power company must have the capacity to ramp up generation in order to meet new demand. This capacity is referred to as SPARE capacity - ie. it's capacity that is online but not being utilized.

    As renewable power is built out, and eventually (maybe) as nuclear power is built out, these power sources will always be preferred over fossil sources if possible due to their incredibly low marginal cost. So even if the renewable power in a region were to double, it is not likely that there will be any spare capacity that is renewable. In fact, as renewables build out, it only increases the amount of spare capacity in natural gas and coal power. As natural gas prices drop, and more natural gas displaces coal, this reduces the spare capacity in natural gas, and increases the spare capacity in coal.

    This is important because only spare capacity can be ramped up to fulfill new demand. So whatever extent the renewable power progresses to, new demand must always be fulfilled by dirty power - and coal power will be preferred if possible.

    This is simply how the real world operates. Analyzing the effectiveness of EV's as carbon mitigation strategies when bounded by the real world of grid power generation and distribution will yield remarkably different results than the "State of Charge" report implies.

    Running on Coal:

    If we go back to the real-world understanding that power companies will always use the lowest-marginal-cost source to fit demand, it's fairly easy to determine what power source will be used to satisfy EV charging. Currently, baseload demand is gradually tamped back each night, and gradually ramped back up each day, to fit an overall peak/off-peak demand curve. If you were to plug in a new, absolutely consistent nightly demand that lasted for ~10 hours/night, then the cheapest way to address that would be for the power companies to tamp down their baseload coal slightly less each night to accommodate slightly higher baseload demand. So slow charging overnight will result in almost exclusively coal-sourced charging energy. Slow charging results in a 10% charging loss - which is factored into the EPA rating, and the USA averages 6% line loss in the distribution of energy from the power plant to the home meter. So if a Nissan Leaf (0.34 kWh/mile) were to be slow charged overnight, then every mile driven in a Nissan Leaf would result in an additional production of 361.7 Wh of coal-sourced electricity. Coal electricity has an average of ~1.1 kg-CO2/kWh, so this works out to a final total carbon intensity of 397.8 g-CO2/mile.

    Running on Natural Gas:

    Natural gas would probably be preferred for "fast" charging (45 minutes), because it's difficult and wasteful for coal plants to ramp up or tamp down very quickly... so even though natural gas is far more expensive on a $/MWh basis, it is far more dispatchable, and therefore is a cheaper option for the power company for fast-response situations. However, when fast charging a battery the charging losses increase to nearly 25% (while the EPA rating assumes slow-charging). So a Nissan Leaf with a fast-charged battery would result in an additional production of 411.0 Wh of electricity. If we assume 610 g-CO2/kWh, then the total carbon intensity is ~250.7 g-CO2/mile.

    In truth, of course, a future grid that has multiple EV's will represent some average carbon intensity between these two points - but what is critical to understand is that increased penetration of renewables or nuclear will not change the above calculation, as only spare dispatchable capacity can respond to new marginal demand. So a Nissan Leaf will have a predictable range of between 251 and 398 g-CO2/mile.

    A Toyota Prius gets 50 mpg combined. According to the Union of Concerned Scientists, the U.S. average life-cycle carbon intensity of a gallon of gasoline (this includes drilling, extracting, transporting, refining, and distributing emissions - as well as the obvious emissions from combustion), is 11.2 kg-CO2/gallon of gasoline. At 50 mpg, the carbon intensity of driving a Prius is 224 g-CO2/mile.

    There is no possible scenario in which driving an EV is as environmentally benign as driving a Prius. Under most real-world scenarios (night-time charging), the actual carbon intensity of driving a Nissan Leaf is nearly DOUBLE that of the Prius.

    Fair scoring:

    However, in the same manner that we can assume a power company will give priority to its electricity production according to price, over a long enough time frame, the oil market should be seen as having a similar priority bias. Therefore, if you imagine removing a few thousand gallons of gasoline demand (by purchasing an EV and shifting that energy demand to coal), over a long enough period the market will adjust to that demand shift by reducing the build-out of the most expensive sources of oil. Right now, that is either tar-sands oil.

    Gasoline refined from the Athabasca tar sands project was scored at a carbon intensity of ~14 kg-CO2/gallon. So if we use THAT figure to compare to EV (we are considering the difference to the environment by selecting one or the other, so if the EV is selected then some number of gallons of tar-sands fuel will not be extracted at a future date), then our per-mile carbon intensity increases to 280 g-CO2/mile.

    If you assume that each vehicle will be driven 12,000 miles/year for 10 years, you can factor in the emissions from the manufacturing of the vehicle. A Prius has a manufacturing carbon intensity of ~9 tons-CO2; and a Nissan Leaf has a manufacturing carbon intensity of ~13 tons-CO2. After 120,000 miles, that works out to 75 g-CO2/mile for the Pruis, and manufacturing 108.3 g-CO2/mile for the Nissan Leaf.

    So the most fair that I can possibly be in comparing the two would have the Prius with an all-in carbon intensity of 355 g-CO2/mile, and the Leaf having an all-in carbon intensity of 359 - 505 g-CO2/mile.

    Jul 09 10:33 PM | Link | 228 Comments
  • Vehicle To Grid Makes Absolutely No Economic Sense Whatsoever.

    A hard sell:

    To begin, we need to determine what the purpose of an electric vehicle is. If that purpose is to operate as a vehicle, then the critical issue must be the versatility in one's ability to move from one place to another. That means maximizing the distance that could be traveled on a moment's notice. For instance: the Leaf can only go 60-70 miles on a full battery, so if someone is called from work to go on a family emergency and then return to work… they probably won't be able to go home. This lack of flexibility from the standpoint of middle-of-the-day options is one of the largest problems facing EV marketing and sales.

    Hence EV advocates are crying out for the ability to plug in their cars at work - not because they want to donate some of that precious stored energy TO the grid, but because they feel anxious from the lack of flexibility in their driving options and want to increase those options, so they demand more energy FROM the grid.

    If the EV owner were participating in V2G, and got some kind of emergency phone call during a time frame when the battery had been discharging energy to the grid - leaving the battery further discharged then when he/she had gotten to work to begin with - then that would further reduce the EV's capacity to function as a vehicle at precisely the time in which it was most critical that the vehicle be able to function as a vehicle. This fact will absolutely reduce enthusiasm for the concept without significant financial incentives.

    The idea behind V2G is that owners will happily allow their vehicle to cycle the battery: fast charging and fast discharging half of the battery while the car remains parked. The impact on the battery life for rapid cycling may not be well characterized, but we know that it is 100% negative - it absolutely will decrease the longevity of the battery, though we don't know how quickly it will decrease the life of the battery. If the battery costs ~$15,000, how much will the power company have to offer the prospective owner of the battery to cut 30% of its lifespan? What about 40% of the lifespan?

    For a Leaf owner, if you assume the natural battery longevity is 10 years, and this V2G insanity will shorten the lifespan to 7 years, we can run the following assumptions:

    The average Leaf owner will show up to work having depleted 30% of their battery from their commute, meaning that they have 20% left to donate back to the grid for half-V2G. That's 4.8 kWh. For reasons that we'll get into later, it's highly unlikely that any given vehicle will charge or discharge energy at a rate greater than ~3 kW from work, and less from home.

    Raw Economics:

    Assuming they contribute 4.8 kWh, recharge 10 kWh, and then again contribute 10 kWh to the grid every weekday, and contribute 12 kWh during the highest demand every weekend day for 7 years, you have the following completely unrealistic scenario where a Leaf owner will donate critical excess capacity during ultra-peak demands on weekdays totaling 27.01 MWh, while drawing a less critical 18.25 MWh from the grid during peak hours on weekdays; and on weekends the total amount donated to the grid during peak periods would be ~8.76 MWh during that 7 years.

    Economically, the price of "Super Peak" (summertime) critical need energy in CAISO can be as high as ~$300/MWh, while the price of the "peak" energy in CAISO averages ~$100-150/MWh, and the price of off-peak nightly energy hovers ~$50/MWh. (These are hub prices, not final consumer prices - the point here is to analyze the potential gain from the perspective of the power companies). CAISO is far and away the market with the greatest profit potential for V2G.

    If we assume the energy drawn overnight has a value of $50/MWh, the energy drawn during non-critical peak hours on weekdays has a value of $100/MWh, the energy contributed during peak hours on weekdays has a value of $300/MWh, during 4 months of the year and the energy contributed during "peak" times on weekdays for the rest of the year has a value of $150/MWh, then assume that the energy contributed during the peak times during the weekend has an average value of $100/MWh… we get a total value of the vehicle's contributions equaling $6,276.00, and a total value of the withdrawals equaling $2701. That leaves a profit of ~$3575.00 for the power company… which seems like a tough profit consideration if they have to convince an EV owner that will lose some use of his/her vehicle and lose 30-40% of the lifespan of a $15,000 battery. But this very sloppy back-of-the-envelope calculation was done assuming that every kWh that was sold for free and every kWh received was contributed for free AND ASSUMING THERE IS NO CHARGING LOSS OR LINE LOSSES.

    Once you factor in a 10% charging loss for all charging overnight, an 8% discharging loss for all discharging, and a 6% line loss for all power sold to the point-of-charge and a 6% line loss for all power received from the point-of-charge… We start getting into real trouble.

    Using the above line loss and charging loss assumptions, now we only have $5427 worth of energy contributed back to the system, and we require $3149 worth of power, for a net profit of only $2278… and we're still sacrificing 30-40% of the longevity of a $15,000 battery.

    Last Mile Transmission considerations:

    The big problem here, however, is last-mile concerns on transmission. It's easy to make big-picture assumptions about the grid as though it had infinite input/output capacity at any point… but that is not the case.

    Most people work in large office towers that have "cubeland" space dedicated to them. If someone in one of those "cubes" decided to bring a dorm-style refrigerator to plug in within their cubical, they would be sharply reprimanded. If 4 people decided to do so on the same floor in the same day, they'd trip the breaker. I have friends who are not allowed to plug in CELL PHONE CHARGERS (~3W) at their cubes due to restrictions on energy draw. This makes sense to people that understand the wiring schematics of the buildings, and know how close any particular breaker runs to the maximum energy transfer level it is allowed. Regardless of how absurd it seems, anecdotes are endless of how a new worker ended up bringing down an entire floor of cubeland by plugging in a coffee pot or radio or something. It's not as though the building had been wired with #30 gage wire or something, but the breaker system had been designed for a certain maximum load, and as more and more desks were squeezed into a space, that load was more quickly approached and surpassed.

    Scale that exact same type of system up, and you get the grid. The power lines running down the street to Doty Scientific Inc have a rated capacity, as do the transformers, and the much smaller lines that connect DSI to the larger lines running down the road. With 4 CNC machines, 10 regular machines, several hundred fluorescent light bulbs, 5 ovens, 2 furnaces, a plug-flow reactor, 1 GC, 2 high-powered magnets in a spectrometer lab, 8 large AC units, 20 computers, and dozens of other low and mid-level power draws, there are many times we may run 50-80 kW (Yes that's a LOT for a small business. Our research is incredibly energy intensive, and we fund that research by manufacturing equipment that is sold internationally.) But even with as much power as we run, a single vehicle being charged at 3 kW would require ~ 6% of our total draw. If we had 5 such vehicles being charged at any given time, we WOULD trip the service limit for the building - something we've never done before.

    A typical small commercial enterprise (say a small clothing store in a small strip mall) might have lights, cash registers, a TV, a fan, and maybe a cell-phone charger for the person behind the register… but have parking for 20+ cars. It's easy to imagine one of these stores might operate comfortably with only an average draw of less than 5 kW. 1 single charging car out front may double their load. If the grid is analogous to a cubeland floor of an office building, then each of these businesses and stores is analogous to a single cube, and each of those EV's might be analogous to something between a personal dorm-style refrigerator.

    In order to adapt to this, the power company will have to re-wire the connection from the buildings to the transmission lines, upgrade the building's service, then upgrade the transmission line capacity for the street and some of the transformers.

    So if your initial reaction to my monetary calculation was "but if we use the cars for frequency regulation (NYSE:FR) they can cycle hundreds of times during an 8 hour period!), consider that the actual maximum power deliver from the battery of a Nissan Leaf is more than 80 KW… then imagine what that would do to last-mile transmission considerations if several dozen vehicles tried to dump 80 kW back through the meter to the local transmission lines on the same street all at once.

    The transmission upgrades in established regions are so expensive that in some cases that power companies have opted for fixed muli-MWh battery stations to accommodate higher loads just to delay upgrading the transmission capacity!

    Think about it.

    Jun 08 1:32 PM | Link | 26 Comments
  • EV's Will Not Reduce Wind Curtailment.

    First, we must understand why wind is curtailed:

    Wind energy is variable, and to some extent unreliable. Due to the extreme ramp times involved in the baseload coal plants, most power companies will not operate wind energy that is greatly in excess of their spare natural gas capacity. If the wind drops, then there has to be enough dispatchable power to cover the load, or you get a brownout or a blackout. This is mitigated somewhat by the distribution of dispatchable loads throughout a region, and certainly throughout a hub on an ISO… but as a region builds out wind power then this will eventually become a concern, and curtailment occurs.

    Furthermore, a power company has to make a call in anticipation of a windy night as to exactly how much they choose to tamp back their coal plants. If the wind is sustained and provides constant power throughout the night, they get very high profits on selling a product that has near zero marginal cost... But if the wind rises and falls, they have to constantly ramp up their peakers near max then tamp them back only to ramp them back up again... All while ramping and tamping their larger CCGT plants as quickly as they can - making the night FAR more expensive than it would have been had they just tamped down the coal plant a little less and pitched the blades on the wind turbines.

    The much lower price of gas has dramatically reduced the instance of nightly curtailment - because they can now plan to have a natural gas power plant running at moderate capacity through the night which can more adequately ramp up or down while leaving the minute-to-minute changes to the peakers.

    In 2012, it's likely that less than 3% of the potential wind energy generation will be curtailed, though as natural gas prices rebound the instance of curtailment is certain to rise again, possibly back into the 10-15% curtailment range.

    But even with today's low NG prices the power companies will pitch the blades and let the energy go in response to minor but rapid changes in wind speed, because it's too much of a headache to try to slam the peakers that rapidly up and down... so they'll use blade pitching to try to get a reasonably constant energy level from the wind turbines once their battery reserves are full.

    This complexity also causes a great deal of daytime curtailment. Since the baseload capacity is typically ramped up near maximum, there again is a narrow range of operational flexibility from the peakers and the natural gas plants… so again there will be some energy discarded by pitching the blades. For instance: in 2010 throughout MISO, nearly 40% of the wind curtailment occurred during peak hours. In 2011, ~58.9% of the wind curtailment occurred during peak hours.

    Another reason for curtailment comes with wind farm design itself: If a power plant builds out a 100 MW wind farm, it's not uncommon for them to only build an ~80 MW (or less) capacity transmission line connecting that wind farm to the grid. The major transmission lines are expensive enough that it isn't cost effective for transmission to be built out with sufficient capacity to handle rare 100% loads. Obviously, during extremely strong winds the blades will have to be pitched back or locked down because the transmission lines wouldn't handle the full load. Due to the very low price of natural gas, this is probably the majority of the issue surrounding wind curtailment today, though as we've discussed before the price of natural gas will not hold, and as the price of natural gas rises, the night-time curtailment of wind will once again occur nightly throughout the wind belt.

    All of the above actions of the power companies are, of course, dictated by the minute-by-minute variation in RT price on the ISO hubs... which is another layer of complexity that I chose to avoid by replacing the ISO hub by a purely rational decision maker - which is fair enough for this post.

    Finally, blades are locked down in winds that exceed the maximum rated wind for that particular turbine, but this represents a very small percentage of current wind curtailment.

    None of this is intended to disparage wind power. As a renewable generation technology that is truly cost competitive with coal on a $/MWh basis, wind has expanded at a tremendous rate and is now generating over 4% of the nation's electricity, and will eventually exceed 20%. But there are complications with wind, and those complications cause curtailments. The higher the penetration of wind, the greater the amount of curtailments that will be required. Understanding why wind is curtailed will help resolve these issues and avoid distractions or fantasies - such as the notion that charging EV's overnight will reduce wind curtailments. They won't. What is needed to reduce wind curtailment is a fast-response demand response. This is often envisioned as some form of grid-to-grid storage, but that narrowing of focus is a distraction. What is critical, however, to reducing curtailment is to have some rapid flexibility of demand that can accommodate a rapidly changing supply.

    The EV charging cycle:

    Plugging in an EV represents an absolutely constant demand increase from ~7:00 pm until ~5:00 am. There's no demand response enabled here... there's just a dependable higher baseload demand. The most profitable response for this will - in all cases - be for the power company to just tamp down their baseload power a little less each night to accommodate a higher baseload demand. (for more on this, see: http://seekingalpha.com/instablog/1005406-glenn-doty/702131-recent-analysis-from-the-union-of-concerned-scientist-is-garbage).

    EV Potential vs Reality:

    If an EV owner had a 3-phase ultra-fast-charger hooked up to a smartgrid controller which could immediately ramp-up or tamp-down the charging rate based on 5-second pricing signals from the local ISO... then you would have an option that would clearly take advantage of the variable energy bursts from the wind and you would directly reduce wind curtailment. But that type of set-up would add ~$5,000-$10,000 to the price of owning an EV, and would still only function during the times in which your EV was hooked up to such a system.

    No current EV owners have such a thing. For current EV owners, and those unwilling to invest another ~$5,000-$10,000 in making them clean, the energy they are powering their cars with is almost certainly coal-sourced.

    Without some kind of ultra-fast charger (which would experience EXTREMELY high charging losses) hooked up to a rapid response smartgrid controller, then an EV is going to be consuming baseload power with absolutely no impact on the rate of wind curtailment, minimization of load following, peaker utilization, optimal capacity utilization, or any other system management concerns or system carbon intensity reduction concerns.

    Jun 06 4:47 PM | Link | 45 Comments
Full index of posts »
Latest Followers

Latest Comments


Posts by Themes
Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.