22 March 2015

Tesla Supercharger

The various standards bodies have determined two levels for the J1772 electric car power supply, which has resulted in no standard for charge rates above that.  Thus implicitly, Level 3 charging means anything above 19.2KW.   There have been three L3 standards to emerge.

These are all loosely speaking "Fast DC" chargers but that oversimplifies what's going on.  The car and charger need to collaborate cooling, current levels and more.  The J1772 connector was really only designed for 10KW.   So something else needed to be done.

The three conventions are TEPCO's CHAdeMO, the J1772 "Combo" connector (aka CCS), and Tesla. Functionally, they are all relatively similar, but they are significantly incompatible.

CCS responds to the weakness of the J1772 connector by adding an adjacent pair of high capacity conductors to the standard J1772 interface.  Much of the J1772 protocol can be retained, and a regular J1772 plug can be connected to a CCS capable car, and it will simply charge at a Level 2 rate.  It's a kludge, but it's compatible.  Several European standards agencies are pushing for its adoption, so while there are not presently many installations, there likely will be soon.

CHAdeMO has much better market penetration.  There are a bunch of chargers all over.  Most of these are on one of the commercial charger networks, although surprisingly many of them are free.  Nissan and several others are compatible, and others, including Tesla, sell an adapter.

The most interesting approach, I think, is Teslas.  They use the same wires as J1772, but they made sure the wiring in the connector and the car itself are much beefier.  Thus the exact same socket can serve both functions.  By being electrically compatible but mechanically incompatible, they can hijack the J1772 wiring for L3 charging. 

Most commercial charging sites have some billing system.  Blink and ChargePoint are two that are popular here in the northwest...both have L2 J1772 and L3 CHAdeMO stations.  But Tesla has gone a different way.  Included in the price of the 85, 85D and P85D is free "supercharging" for life, and you can buy this feature for $2.5K for the 60.   Unless you do a tremendous amount of long distance commuting, this is not too cost effective.  But the high price is helping Tesla afford to build out the network of superchargers.  As of this writing, Tesla has 403 stations around the world, almost half in the US.  Many major routes are supported, but there are still a great many gaps, and for the time being, they have not built many close to cities, figuring that home charging will serve most of that need.  (But that leaves gaps.  For example, if all you have is L1 charging at "home" (e.g. a hotel or billet) but you have more than 70 miles a day of errands to run)

Last summer, in the leadup to the announcement of the BMW i8 and i3, Elon Musk widely dropped the hint that he was negotiating with BMW for them to be compatible with the Tesla supercharger.  This would have been a terrific deal for both sides: BMW would get a widely installed base of fast chargers, and Tesla would have a partner for the very expensive and time consuming buildout of their network.  For the time being, Tesla is being very generous with their installations, basically funding it from the high prices they can charge early adopters.   But BMW made it clear that they would not participate in this and now that the i8 is available, it uses the CCS "Combo" connector. It's unclear whether this was because of pressure from EU politicos or what.

Tesla presently charges once for lifetime supercharger use, but there's no particular reason they need to.  All the existing cars will continue to get free charging, but eventually they'll go away.  If they want to bill, either by KW or by the month or something, it's easy to do.  The car can identify itself to the charger and send a bill to the appropriate customer.  This could easily be implemented over the "proximity" conductor, or even by some sort of RFID.  This is exactly what the existing services, including Blink and ChargePoint, do.  If Tesla is clever about it, they can sell a fast charging adapter that includes a supercharger license and allows you to plug your CCS or CHAdeMO vehicle into the supercharger.  I can even imagine supercharger stations offering single use rentals of these gizmos.

If one of the other plug formats ends up winning, Tesla has a simple solution.  They already have a CHAdeMO adapter, and they've said that once more than a handful of CCS stations are on line, they'll develop an adapter for those too.   The stations themselves can be gradually converted over as demand moves.  Most of the cost of these stations is in the real estate and preparation, including the wiring, which is the same for all.  The charger and connector is probably less than 10%, and I bet even most of that can be switched over easily.  Although if this happens, I suspect most Tesla stations will be sold to some other franchisor. 

14 March 2015

Monocausality

Much of the bad thinking in the world come from monocausality: thinking that every event just has a single cause.  Once in a while, this happens, but it's pretty rare.

Conservatives tell us that if we raise the minimum wage, unemployment will spike and the economy will decline--because if the price for employers goes up, demand will go down.  Yet, in literally thousands of practical experiments, not only has this never happened, but in quite a few of them, the economy and employment situations have actually improved significantly.  What must be happening is that some other factor--for example, perhaps, the extra disposable income of low wage workers that the hike provides--is stimulating the economy enough to overcome the downward pressure on low wage employment that the wage hike is causing.

Another popular example is the claim that the reparations after WWI caused WWII.  While this is certainly a large factor, there are others.  A big one that doesn't get much recognition is that large german businesses allied themselves with Hitler when they realized he intended to create a large mass of extremely cheap workers and violate the weapons limitations agreed to at Versailles.  Hitler's policies were great for them...for a little while.  There were lots of other factors too, of course.

Another example is that it is claimed that printing lots of "fiat currency" will surely cause hyperinflation.  It did in Weimar Germany, Zimbabwe, Argentina, etc.  But it didn't in the US during either the great depression or the recent "great recession."  It hasn't in Europe either.  Non economists and "classicals" insist there must be something else going on, such as the currency being un-backed, for example by a Gold Standard.  But this also has near perfect track record of failure.   Keynes explained this:  It's not the backing or the ultimate supply of currency, it's demand relative to supply.  If there's very high demand for currency but not for goods and services, the value of the money will stay high while accompanied with little inflation or even deflation.  Those who can, will hoard currency.  In the recent recession, this is exactly what happened.  Investors suddenly flipped from being hot on subprime and other high debt instruments, especially those based on real estate, to being terrified of them...including the ones they'd bet on heavily, and wanted to sell them and move into cash.

Another example is the so called Laffer Curve.  The concept predates Arthur Laffer by at least a century but he advised the Reagan administration when it became very popular.  Simply put, it's that taxes put a downward influence on the economy.  This is surely true, but the empirical data that we have tells us the almost exact opposite.  Reduced taxation almost always correlates with downturns and increased with upticks.  What's probably happening is that the spending that's done with the tax money is significantly more stimulative than the taxation is de-stimulative.  For most people and business, this seems to go right up to 70% taxation and higher...the 91% marginal tax rate they were paying didn't cause Howard Hughes or J Paul Getty or Henry Kaiser to pull out of business.  The Laffer curve may well be correct, but conservatives seem to only look at the part to the right of the peak...the part that would justify their tax cuts...but we're not dealing with that part.  To left of the peak says that taxes should be higher.

There are lots more examples.  The causes of these events are numerous and interact in complex ways.  In mathematics, we call these sort of equations "equilibrium" equations and they are generally not trivially solvable and tend to end up having several solutions or a solution which itself is a fairly complex equation.   Simplifying models, such as Hick's IS-LM and the Phillips and Laffer Curves in economics can be very useful--provided you actually look at these curves correctly and understand them and their limitations.  They're not a perfect answer by any means, but trying to solve these complex problems by ignoring important--perhaps dominant--factors is unhelpfu.

added 26 Apr 2015
Greg Mankiw argues here , astonishingly, that there is only one pertinent point of the proposed trans pacific trade agreement and that is that freer trade is is better.  What if, for example, the freer trade causes millions of job losses (through, for example, outsourcing) and the new trade only generates a fraction as many new opportunities?  What if the freer trade grants patent violators the right to sell cheaper products competing directly against those that funded the R&D with impunity?  Of all people, Mankiw should understand that nothing is ever so simple.

12 March 2015

Gehry, Gaudi, and Logic

Two of the supposed greats in modern architecture live by the dictum:

"There are no straight lines or sharp corners in nature.  Therefore, buildings should not have straight lines or sharp corners."  -Antoni Gaudi

Everything about this statement is wrong.   Lots of stuff in nature is straight and/or sharp.  e.g.:


More importantly Gaudi's syllogism is missing its major premise, which I believe is "Architecture must mimic forms found in nature."  Why?  That's plainly not a requirement for many of the buildings that most of us think are beautiful.  The Eiffel Tower is one of the most beautiful structures in the world.  It has lots of curves, but they are smooth curves, which are fairly rare in nature, and lots of the straight lines and sharp corners that Gaudi so detests. 

Frank Gehry seems to have been cut from the same mold--his theory is "deconstructivism", which is about breaking up the straight lines and conventional components of a building to make way for new arrangements and ideas.  A few deconstructivist buildings are not complete disasters: the Sydney Opera House and the Seattle Public Library are among them.  They are eye catching and fairly functional, even if they are not the most attractive. But Gehry has taken this to an extreme that I find appalling.  I have nothing against experimentation, even when it is jarring.  Once in a while, a good new thing comes of it.  But I don't wish to be jarred over and over again by the same failed experiment.   Gehry's buildings are ugly and stupid looking, every one.  I wish people would stop giving him money.

Gaudi designed some nice buildings in the early part of his career, but he seems to have lost his mind about 1900...and the illogic above explains it.