Wednesday, December 07, 2016

Batt-decided!

A multi-pronged pack design strategy paid off, and in the end we didn't get any finished CAD drawings, but a simple re-stacking of blocks to test all options produced the simplest (and hopefully best) approach.  The 88 buddy-paired LEAF modules will fit fine in 'pancake stacks' with none of the more complex 'library' orientation that requires heavier custom bracing.  The flat stack is easiest to build, but poses some thermal risk, as the topmost module will experience the most heat while the bottom one could get cold.  We believe this risk to be insignificant because of several factors.

  1. The vehicle's duty cycle.   The drive in a Ranger is of a very similar power level (84 kW vs 80 kW in a LEAF), but since our modules are doubled up, they will be discharging at roughly half the amperage as their original application, generating less heat.  Likewise, they'll also go through a slower Level 2 charge, generating less heat.  The Ranger EV will not be subject to DC Fast Charging, which poses the greatest heating concern.  A side benefit of slower-rate charging is that the truck will only be able to do one full charge-discharge cycle per day, giving plenty of time for cooldown.  
  2. Our mild climate.  The Puget Sound Region is ideal EV territory, with moderate summers and mild winters that in general keep battery packs happy without supplemental heating or cooling.   As the LEAF modules are particularly vulnerable to heat, there is extra insurance in that mileage will be lower during summer when school is not in session. 
  3. The insulated box.  What looked like a simple fiberglass battery box is actually a wood-cored composite on the bottom and sides, providing strength and insulation from the cold, while allowing excess heat to radiate through the lid.  The lid has two passive vents that could be upgraded with forced air if temperature issues arise. 
  4.  Testimony from other LEAF module users shows that in an unventilated Library stack (or row of modules), it is the center module that gets hottest, though not by a troublesome margin. That's info from the most similar truck conversion that's been built; http://www.evalbum.com/2898
Here's the current status.  The original base plates for four stacks were cut in half for use with two stacks, which we're through-bolting to the battery box, and then the stacks of modules bolt onto their factory base plates with factory top plates.  We've laid out all electrical interconnects with tape, and are bolting the stacks together for proper compression to 1.333"/module before adding busbars.  The multiple stacks allow student work at low voltages on the individual stacks, while I'll make the higher voltage interconnections and final high voltage pack connections.

Monday, December 05, 2016

A DC Fast Charging Death Spiral?

I've tried to inform GM that myself and others will have no interest in the Bolt until they announce and demonstrate a commitment to growing DC Fast Charging (DCFC) infrastructure, but to no surprise, there's no reaction from the General.  With the market switching over to long-range EVs mixed with short range PHEVs that don't have DCFC, the market share of medium range EVs like the LEAF, SOUL, Focus and i-MiEV will most likely decline, even as these medium range EVs increase in absolute numbers.

Contrary to the expectation that new long-range big-battery cars will result in more DCFC utilization, I posit that it could mean less, as DCFC will mainly be needed by the long range cars for turnaround at their final destinations (ie: in Portland and Seattle, but not so much in-between).  Local region drivers that might've stopped for a midday DCFC will no longer need to, and the longer highway trips are rare.  Rare enough to not make a viable business proposition of the silly subscription model that AeroVironment uses, and since any CHAdeMO or CCS- equipped car can't compare favorably to Supercharging, the share of trips beyond 180-200 miles done by non-TESLA evs will remain tiny.  Spending at least an hour on a 40 to 50 kW DCFC station before your next 180 mile segment is significantly slower than my usual road trip pace, and less than half as appealing as a TESLA supercharger, even if 'only' on a 60 kWh Model S (which matches the Bolt EV capacity).

The Puget Sound region has never been adequately served with DCFC, and the situation is only getting worse.  The initial DCFC deployment was bad design, with the Dept of Energy's 'donut hole' exacerbated by Blink's bankruptcy and worsened further by Car Charging Group's inability to maintain the small number of stations that were actually deployed.   AeroVironment did a decent job of station maintenance, but hasn't added a single DCFC in almost four years, and reliability seems to be slipping, with the critical Tumwater station down for weeks now due to a worn-out handle.  Thrice in the past two weeks I have wanted to make a regional EV trip requiring only one DCFC before turning around, but most of the DCFC stations were offline, with the remaining couple of options perpetually busy.  Dealers that host DCFC have proven to be poor managers of that resource, never posting to Plugshare and rarely accomplishing repairs in a timely manner. In my experience, the worst of these are Car Pros Kia in Tacoma and Olympia Nissan.  Puyallup Nissan is slightly better, but they let the issue of a simple gland nut reconnection go unattended for about two years, increasing the risk of wear, downtime, and shock hazard.  Puyallup Nissan at least ensures 24 hr access and fair pricing for all comers at $10/hr, but has been increasingly unavailable due to a seemingly minor thermal sensor error, with no communication about a repair schedule or repair attempts.  Since Olympia Nissan and Olympia Chrysler are tied at the hip, I've informed them that their failure to maintain DCFC and even ensure access to the Level 2 chargers means that I will not purchase a Chrysler Pacifica PHEV there.

I've been forgiving of EVGO's less-than stellar reliability due to their generally prompt repairs, but now I suspect subpar equipment, as the same problems are recurring regularly.  Early shutdown, about 2 minutes after starting a session is worse than not starting at all, as that's just enough time to head into the mall, with travel plans killed after 20 minutes of shopping to find an uncharged car and the next user waiting in line....  Poor reliability and extra high pricing are a business plan destined for failure.

Back to my premise; if the number of DCFC stations in the Puget Sound area remains small and their reliability continues to worsen, the addition of even a small number of professional DCFC campers in the form of Lyft drivers with GM Bolts could crash the fragile system.  The Lyft experiment will fail, and us medium-range EV drivers will be relegated back to our home garage and daily commute.

Get on the ball GM- announce support for your product, or you'll take down not just the Bolt, but also your Frankenplug cronies; Ford and the Germans (if not the entire non-TESLA EV market), by giving the system a nasty jolt just as what minimal Federal support there has been gets Trumped...