My brother drives an S and his analysis is that here, relatively mild compared to your situation, it is much more efficient to use the seat heaters than heat the whole cabin.
Some engineer on YouTube did the calculations and after ten years of nothing but supercharging the capacity had dropped 30%, probably not enough to replace (yet).
As electric power trains for transport become widespread they need the grid to be adapted to deliver the electricity where it is needed. The corollary of this is that the infrastructure for supporting internal combustion engines will become progressively redundant. Refineries, Fuel storage and distribution, pipelines, road tankers, filling station with huge underground tanks. All that represents a lot of assets and land value that could be liberated. It is a fuel refining and distribution network that comes with a high operational cost.
We have been here before. When internal combustion engines replaced horse power, the network of stabling, feed distribution and dealing with the vast quantities of manure became redundant. That was a far bigger change than the challenge presented by EVs today. I am sure those early motorists worried about where to get fuel and repairs in a society still oriented towards the horse and buggy.
All the roads and vehicle servicing infrastructure is in place, easily adapted to EVs. There is also a mature electricity grid and the indications are that this can be adapted to deliver the power where recharging is required. It is not building from new, but it will have to adapted. I am sure there are electrical engineers worldwide who are very keen to take on the challenge.
I’m already seeing roadside charging stations appearing in my city. When it comes to a lamp post near where people live, urban motorists without home chargers will see a green light for the next car to be an EV.
Whether patterns of use will change… Best of luck predicting human behaviour. Driverless cars are another thing. That will take some time if the progress towards driverless trains is any indication. Despite railways being a highly controlled environment with well developed signalling systems, there is lot of resistance towards driverless operation. That is not a technology problem. I don’t doubt that driverless vehicles will be a hugely disruptive technology leading to a lot of changes in behaviour that will take the established motorist by surprise. It won’t be all bad. Being taken home in driverless car summoned at any time of the night or day is a big plus for every sort of person. But it will be the less personal business of moving stuff around in trucks over long distances on fixed routes that is the easiest part of the problem to solve. Driverless technology has a long way to go.
Bolding mine. Are normal temps going to make you want to use AC? I wonder how much power that will use compared to heat. I’m thinking it would use more.
AC use taxes the battery much less than the heater on the model 3 does, since i believe the heater only uses electrical resistance elements, rather than a heat pump (more efficient) or getting it for free (waste heat from the engine) like an internal combustion vehicle does.
Air conditioning typically has a Coefficient Of Performance of ~3, so in general its about 3X more efficient to cool with AC than it is to use electrical resistance heat.
Also note that EVs in general should be more efficient at using AC than internal combustion vehicles are, since an IC has to take mechanical power form the engine, convert it to electric current using the alternator (with some loss there), then drive the compressor using the electric current.
With an EV you just go straight from battery to the compressor, with no extra steps.
Not only will cars mostly be charged overnight at off-peak times, but electric cars synergize very well with wind power, which is probably going to be an increasing share of our power generation in the near future.
This is correct, but the difference will be less if/when they switch to using heat pumps for heating as well. The newest Leaf uses a heat pump, and I see significantly less power consumption than with the old resistive heater.
But even if power consumption were equal, AC will still have less impact on range because the batteries are more efficient at ambient temperatures requiring cooling than temperatures requiring heat.
Even in northern Alabama, our Volt gets less battery range in winter than summer. Something like 35 miles on coldest days of winter, 38 miles in summer and up to 44 miles in spring/fall.
I think part of the reason is, in winter, the car uses a lot of power just keeping the battery packs warm.
It helps if you “pre-condition” the car before leaving home - that way you are heating up the car (including the battery) with power from the outlet rather than the battery.
Sorry, that should have read “better range in summer than winter”. The numbers are correct.
Huh. Color me educated about heat vs AC in electric cars. Interesting. Thanks.
There was no realistic chance that any new technology would replace lithium-ion batteries in the last ten years, as there was nothing then on the horizon of feasibility and it takes about that long to bring a complicated technology to market, notwithstanding the extraordinary cost of building a facility to manufacture a product with the contamination requirements of a modern lithium-ion battery. However, there are technologies on the cusp of feasibility which offer to double or more the energy density and specific power of lithium-ion batteries: lithium-sulfur (Li-S) and sodium ion (Na-ion), both of which offer a potential material cost savings with deeper discharge, higher C-rate, and/or greater cycle life. The degree of improvement available for lithium-ion is pretty marginal; the incremental improvements have been in getting better quality control and finer size of the graphite anode, and tailoring battery chemistry to specific types of applications (highest specific power vs. most level current vs. deeper discharge vs. high cycle life), but there is no opportunity to improve any of those characteristics by anything close to a factor of 2 or better.
For “microgrids” (an installation supporting a single household) batteries are fine because the power demand can be managed. But as anyone who has maintained an off-grid system will tell you, you make a lot of compromises to stay within capability, or spend a lot of money to support a high peak usage level. Batteries are certainly not cheaper for large regional or national grids, because there is no practical way for such grids to manage demand without threatening brownouts. Having enough storage capacity to store energy to load balance across an extreme demand cycle (mid-summer or mid-winter, depending on clime) combined with being about to provide the specific power to meet peak energy demand requires a massive amount of excess capability compared to typical usage. Nor is there sufficient manufacturing capability to just produce that amount of batteries for large grid storage and load-balancing, and building up that capability would require huge capital investments.
On the other had, we already have natural gas plants or the ability to convert other types of plants to use natural gas, and natural gas power plants are one of the cheapest types of plants to build on a capacity-weighted average basis. In terms of energy storage, hydroelectric is cheaper but has other limitations (how much and how fast water can be released, as well as the environmental impact of building large, variable level hydroelectric reservoirs), so the easiest way to load-balance a large grid–which is crucial for maintaining a reliable regional or national power grid–is by adding natural gas plants which can be spun up rapidly to meet demand. Natural gas is a hydrocarbon fuel (although less CO[SUB]2[/SUB] emitting than coal) with vast known reserves, the ability to be produced from shale or coal (albeit at the cost of a higher carbon footprint and more environmental impact), and natural gas plants could be readily converted to combust methane or methane-DME mixtures produced from biomass or other renewable sources.
Charging electric vehicles at night makes sense from a conventional power production and infrastructure standpoint where power comes from relatively inflexible coal and nuclear power generation, but if the desire is to use renewable solar energy directly, it makes more sense to charge them during the day when peak power production is occurring. It also makes sense to reduce the size of batteries to a minimum so as to minimize excessive cost for unnecessary range beyond that needed for typical commuting. The confluence of autonomous vehicles and high density battery storage for electric vehicles means that they can support commuter applications, then go off and charge during non-peak usage but during peak solar power production, preferably somewhere remote that doesn’t require the poor utilization of real estate in the form of vast parking lots in front of every commercial and retail building which are frequently empty and do nothing except contribute to the “heat island” effect and prevent effective precipitation runoff.
Estimates about wind power utilization vary widely, but there is general agreement that while there is an enormous amount of potential energy available in atmospheric dynamic energy, actually tapping into it for constant utilization is very difficult. Except for offshore wind farms that can rely upon predictable wind patterns, power from wind is highly variable and only regionally available. Onshore wind farms also have the problem that no one really wants to live near one (if you’ve spent any time near a wind farm with its constant ominous oscillatory pumping and the auditory effects of slightly out-of-sync turbines producing a headache-inducing buzz you’ll appreciate), and there are real concerns about their effects on pollination and bird and insect migration, as well as the very large footprint required per kWh produced. Wind power production has grown at an exponential rate over the last decade, but it is still a tiny fraction of power production and its ultimate scaleability to meet expectations of 20% or more is dubious at best.
Solar power has the greatest potential of a truly renewable, sustainable power source, but there are some massive capital investments that have to be made in an power storage and distribution infrastructure that are unique to its specific mode of generation. Varun Sivaram’s Taming The Sun is an excellent in-depth look at what is required from an infrastructure improvement and capital investment basis to make widespread solar power a viable source to replace conventional hydrocarbon sources energy. For the foreseeable future solar is just a part (although hopefully a growing part) of the national and global energy portfolio that includes natural gas/combined cycle, 2nd generation biomass liquid or pressurized gas, nuclear fission (although hopefully moving beyond 3rd generation boiling water and pressurized water reactors with an incredibly wasteful once-through fuel cycle), and ultimately culminating in controlled nuclear fusion and/or methods to convert remotely-generated solar power into storable and fungible forms of potential energy that can be transported to point-of-use, e.g. synthesized fuels with a carbon-neutral production basis.
To address the question of the o.p., the US grid (which is actually a patchwork of regional grids) is certainly in dire need of a complete overhaul, not just to support an increased demand for electrical power for transportation but also to protect it against both malicious threats (cyberwarfare) and natural hazards (extreme climate effects, coronal mass ejection causing a geomagnetic storm like the 1859 Carrington Event). The cost of this is almost incalculable–certainly on the order of a trillion dollars–but necessary to ensure that the US is able to maintain itself economically and technologically, the future of electric vehicles notwithstanding.
Stranger
One of the issues in recharging electric vehicles was the non-standardized chargers needed for different makes and models of cars. Have chargers been standardized?
I believe all current/recent electric cars sold in the US have J1772 for level-2 charging. Tesla doesn’t have a separate J1772 charge port, but they include an adapter for it. Level-2 is basically for overnight or all-day (workplace) charging, rather than while-you-wait charging. The exact charging speed depends on the installation, usually 240V 20~40A, some of them allow 80A.
The issue is fast charging, which is something you only need for out-of-town road trips. There are 3 major competing standards: Tesla Supercharger, CHAdeMO and CCS. CCS seems to be getting more popular - used by GM, BMW (at least in the US), and now Kia and Hyundai. The Tesla Supercharger standard is currently only used by Tesla. CHAdeMO is used by most Japanese manufactures (notably the Nissan Leaf).
Admittedly being Jon Snow/Sgt Schultz about any technical details, and you apparently having some deep knowledge, I was wondering what you thought of this article - which suggests building a ‘bottom up’ grid: I have solar panels, and belong to a microgrid of my neighbors. If I need more power, I request it from my microgrid. That grid in turn is connected to other neighborhoods, which share power. If that level is short, it bids/requests up, and so on (turtles all the way up) until eventually you hit large scale hydro/nuclear/gas/coal (and apparently those resource are already tied at least loosely into grids).
One of the ideas in the scheme is that computer management of the information needed to management that power is relatively cheap, so you harness that.
Which is why they synergize well with electric vehicles. If a car needs six hours to charge but it’s parked for twelve hours, then you can spend only half the time charging. If the electric company makes power cheaper when the wind is blowing, and more expensive during lulls, and the customer has a smart meter, then you can use the wind power when it’s being produced without being hurt by the times it’s not being produced.
EDIT: Folacin, that’s precisely what an electrical grid already is. But it’s not a complete solution, because sometimes the wind and/or sun are low-producing everywhere at once (or at least, the total production is significantly below average), and coal and nuclear plants can’t ramp up quickly enough to meet the demands of such a situation. Natural gas plants can respond quicker, which is part of why they’re currently popular (the other part being that fracking has made gas very cheap).
I understand that much :). It’s probably outside the scope of this topic, but the article deals with organizing the grid to handle all of the potential sub-utility scale power generation that is coming.
The problem isn’t the management of information but actually having the physical capacity to produce enough power for peak demand, and to ramp up power generation quickly enough to meet demand without causing brownouts. Rapid changes in demand are not only a problem in making energy available but also maintaining stability of the distribution grid. On a small scale this can be done just by overbuilding to meet peak demand and/or limiting delivery via some managed process, but on the scale of a regional power production and distribution system where consumers expect power to be available upon demand regardless of what the overall load on the system is.
Building a regional-scale grid from the bottom up would require a large amount of standardization, not only in how the grids interface but in how they are managed internally. I don’t think it is wholly infeasible but it would represent such a radical shift in how electrical power production is managed today that it would essentially be a completely new system, albeit one that might be far more robust against damage or failure if built using a robust architecture (e.g. cybersecurity and authentication built into every aspect of the system from production to end user). It is easy to conceive of such an infrastructure in diagram, but actually realizing it is far more challenging. It is comparable (but physically more complicated) than a dynamic peer-to-peer distributed global computer network to replace the Internet, and yet, we’re still wedded to TCP/IP protocols rooted in the 'Seventies with its inherent backbone physical infrastructure (which, like our electrical distribution infrastructure, is a loose organization of individually owned networks) with no truly dynamic P2P network that functions independent of the transport and Internet protocol.
The problem with variability it two-fold; one is that it makes power distribution networks difficult to manage; if you are generating power in excess of what is is demanded you have to store or dump it in some way, and if demand exceeds supply you either have to rapidly spin up additional power generation or start preemptively shutting down areas of the grid to prevent cascading failures.
I know that a lot of people think that a dynamic pricing structure will allow some kind of “invisible hand” regulation of the grid to make it stable, but the reality is that residential consumers do not monitor their hour-by-hour usage and commercial users are often inflexible because they have to maintain work hours, keep customers comfortable, maintain production, et cetera. Wind power that provides a large peak one day and nothing the next is not an opportunity for an electrical utility provider; it is a problem. Worse yet, in a system where energy is sold at unregulated (market-based) prices, producing excess power can drive prices literally through the floor, where power generators literally cannot make a profit or even enough to cover overhead, and yet have to keep operating the plant because it cannot spin down or up quickly enough to meet the non-peak base load once the transient power source goes away.
This is not a hypothetical scenario; it actually happened in Chile, where investment in “merchant plant” solar (with price based upon the market rather than a long term purchase agreement) combined with reduction in demand for copper which was one of the major users of peak electrical power, which resulted in price for power hitting zero in the afternoon throughout much of 2015 and 2016. Such “market flooding” is a problem for any source of power where the rate of generation cannot be controlled and can fluctuate unpredictably, or where forecasts based upon projected demand overpredict the needed capacity. Solar has this problem; even though solar power generation can be reliably predicted in many areas such as the Atacama Desert with its high altitude and very limited cloud cover, having radical daily transients without some way to practically store and redeliver power upon demand, or just having overcapacity at peak generation times is enormously problematic.
Wind is even worse because onshore wind turbine generation just isn’t reliable enough to make even weekly projections, it has a massive footprint per specific energy capability, and it sees radical transient production throughout the day/night cycle. Hoping to load balance that with electric cars charging at peak generation times is not a very good strategy for stabilizing the power grid. If we actually had sufficient wind generation capacity to power a sizable majority of hypothetical electric vehicles sufficient to replace consumer and OTR applications for today, we’d also have to have some alternative means of producing power when wind generation underperformed compared to required demand.
In other words, our electrical generation and distribution system cannot be built to supply the average baseload, or even the peak load averaged over a month; it has to be able to supply up to the absolute peak demand, or else we have to somehow triage which customers get priority. The system we have today in the United States (and with caveats, also in most of the developed world) was developed in a somewhat ad hoc fashion where government regulation of pricing and a predictable constancy of electricity generation capacity created a system that could meet those demands including peak fluctuations without concerns about being able to make a return on investment. A system that has more fluctuation in when and how much power is generated without some way to store and then deliver rapidly enough to match peak demand is not indefinitely scaleable and is going to be inherently destabilizing, both fiscally and physically.
Stranger
Thank you Stranger for your (once again!) very, very informative posts.
Colorado has a $50 EV registration surcharge, but you get a cool sticker. I like that more than a mileage based fee, but I expect that to happen when EVs reach some critical mass.
I find it is easy to keep the passengers warm with the seat heater and jackets, the difficult thing to keep warm is the front windshield. For example, looking at my log, the same drive on a cold, clear day may use 210 Wh/mile, but add in snow or freezing rain and it is up to 445 Wh/mile. Most of that difference is running the front defroster, and is even worse, as speed is lower (using less power) on a snowy day than a clear day.
Agreed in that instant pricing won’t change when I turn on the oven or even the washing machine, but it will change when the car charges. If there is instant pricing, then the natural thing to follow is a setting in the car to only convenience charge when the price drops below a certain amount. So, if someplace to sink excess capacity is needed at that moment, cars and home batteries will come online and soak some of it up. Their penetration at the moment may not be high enough to make much difference, but this thread is about what will happen in the future when they are much more common.
Looking at the Chargepoint map, across several charging networks Boulder, a relatively small city, has over 200 public charging locations. What would happen if they were all used at once? I have no way to know, and that probably isn’t anything near the density needed to support 25% of cars being electric. Clearly though, that is an installed capacity that the power company thinks it can support. Or they don’t care.
Every vehicle I have owned appeared to have the AC compressor direct driven by a belt off the engine. The only electric in it was the clutch. Has something changed in the past 5 years that I haven’t investigated? Or did I misinterpret what I thought I saw?