Wednesday, August 26, 2009

Greenhouse Gas Emissions from One Gallon Challenge

Just to complete the previous post, here a map of the United States describing the fuel mix used for electricity generation. Source: EPA eGrid page.

Emissions from tailpipe.

Gasoline and hybrid-electric cars emit through their tailpipes. They produce emissions and reduce air quality in the urban centers where population density is high. An electric car does not have a tailpipe. Here is a chart of CO2 directly emitted by the cars during the One Gallon Challenge. [Based on DOE data, more info at Carbon Dioxide FAQ]

Emissions from well to wheels.

Even though an electric car does not have a tailpipe, an electric car produces greenhouse gas emissions where electricity is generated. These emissions are usually counted as "well to wheel" emissions, that is, a full life cycle of fossil fuels from their extraction to actual process of burning. They are differentiated from "tank to wheel" (or "plug to wheel" for electricity) emissions from the tailpipe.

Gasoline-powered cars, both conventional and hybrids, also produce upstream emissions. Most significant ones are emissions from the oil refineries that convert barrels of oil in gallons of gasoline. [Based on Argonne National Lab GREET]

A few comments:
  1. Since we kept track of when we charged the car, we know that natural gas was used as a fuel to produce electricity for our drive. In general, you do not know which fuel was used to produce electricity for your air conditioner or flat screen TV. A common approach is to take US average grid fuel mix with 50% coal, which is almost twice as polluting as New England grid.
  2. The 75 MPGe well-to-wheel number quoted in the AutoBlog was based on the US average grid. Had we taken New England numbers, we'd get above 100 MPGe. Had we charged our car with wind or solar power -- no fossil fuels used -- we would have infinite MPGe!!!
As you can see, an electric car is two and a half times cleaner than a gasoline-powered Prius with its strong 52.9 MPG. The main difference is the fuel. Our Porsche effectively used natural gas while Prius used gasoline. Had Prius used compressed natural gas (CNG), its greenhouse gas emissions would be probably about the same or even slightly lower than those from an electric car.

A conventional engine CNG vehicle, for example, the only mass-produced available in the states Honda Civic GX would not be able to rival an electric car because it lacks a hybrid option. There have been prototypes of CNG Hybrids, for example, Toyota Camry CNG Hybrid and Opel/(Saturn) Astra Caravan CNG Hybrid, demonstrated in the past year.

CNG Hybrids or Electric Cars?
As a society, should we focus on electric cars or CNG Hybrids?

Advantages of CNG Hybrids.
  1. Price. Electric cars are quite expensive due to the cost of batteries. When our Porsche was built a few years ago, the battery cost for ~100 miles range was $36,000. If the car were mass produced today by a major automotive manufacturer, the battery cost would be closer to $15,000. It is still much more expensive than a hybrid powertrain.
Disadvantages of CNG Hybrids.
  1. Leakage. Natural gas leaked from the refueling stations and from the car itself due to leaky hoses is a potent greenhouse gas, four times as damaging as CO2. Because of that, developing a complex infrastructure for natural gas refueling may end up producing just as much harm to the climate as today's gasoline cars.
  2. Renewable energy. Electric grid in the US will become greener with time. Emissions from the grid are already tightly controlled and budgeted (see RGGI, an initiative by ten Northeast and mid-Atlantic states). Addition of renewable sources, such as wind energy or solar power, to the fossil fuel powered grid will inevitably reduce well-to-wheel emissions from electric cars.
    An average car in the US lives for about 15-17 years. An electric car sold TODAY emits as much greenhouse gases as TODAY's CNG Hybrid car. In ten years, an aging CNG Hybrid car will emit the same amount of CO2 per mile or even more, due to leaky hoses. In ten years, the same electric car will emit less CO2 per mile because its electricity will be cleaner. Electric cars are future-proof. That is one of the reasons our team is the Electric Vehicle team.

Tuesday, August 25, 2009

Gallons in an electric car?

MPGe and gallons equivalent is a convenient metric to compare electric and conventional cars, but it's pretty evident that an electric vehicle does not burn gasoline. What does an electric car really use?

We monitored the charging behavior of the car closely and matched this information with the hourly data from ISO New England (wikipedia: about).
Here is a plot of electricity demand in New England with the profile of car charging overlayed. Both axes are in the units of power, MW stands for a megawatt (1 million Watts), and kW is 1000 Watts. An average incandescent light bulb is 60 Watts, an average compact fluorescent bulb is 20 Watts, and an average car has about 160 horsepower, equal to 120 kW. That makes 6,000 CFL bulbs in an average car.

The electricity demand follows a fairly typical summer profile. We use much more electricity in the afternoon (peak demand) than at night. The main use during the day is air conditioning.

The car charges whenever it is plugged in. Most electric cars will be charged at night in garages, on driveways, or on public charging spots. Just like we charged at our friendly hosts, Ford of Greenfield. We also charged during the day, in fact, during peak demand. Did it matter for the utility company? We'll see shortly.

United States fuel mix for electricity generation

United States electricity grid consumes about 50% of coal. Electricity produced with coal is particularly "dirty". Yet all states are quite different. Here is a fuel mix used in the New England: coal is only 11%, and renewables (including hydro) is greater than that. As you can see, the electricity demand on August 19th and August 20th were less than maximum capacity, so we did not run the dirty diesel generators.

The car charges "on top" of the demand bar, so we used capacity from a combined cycle natural gas plant. For this race, our Porsche was actually powered by the natural gas. It is a fossil fuel, yet its source is all domestic.

We were concerned whether our electric car mattered for the utility since it was being charged at peak. For the comparison, we looked at the difference between two consecutive days, Wednesday and Thursday. Wednesday was slightly hotter, so consumers used a bit more electricity. A difference between Wednesday peak and Thursday peak is 871 MW sufficient to recharge 414,762 electric cars like our Porsche

We will continue with the summary of CO2 emissions from the race.

One Gallon Challenge: 164 MPGe

Last week, the Porsche participated in the One Gallon Challenge, part of Boston Greenfest 2009. We traveled from Greenfield, MA to Boston City Plaza using 23 kWh of battery power, equivalent energy to 0.65 gallons of regular ethanol-containing gasoline. That translated to 164 miles per gallon equivalent (MPGe). Here is a link to our path. We also got lost on the way which added a detour, so the total length of the trip was 118 miles.

View Larger Map

Both in Greenfield and in Boston, we had a tent for the car and a lot of visitors. Outreach and public education is an important part of our effort. For everyone who stopped by to see our car last week, thank you.

A day before the race, we drove from Cambridge, MA to Greenfield, MA. We completed this drive on a single charge during the prior night. In Greenfield, MA we charged using 240V AC at Ford of Greenfield which generously let us park there at night.

However, we were not able to fully charge during night due to logistical difficulties, so we started the race with 80% full batteries. We had to recharge on the way. Luckily, Colonial Dodge of Hudson, MA let us park next to their service bays during the day and fill up our battery.

Charging at Colonial Dodge

We also had two cars, a hybrid-electric Toyota Prius and conventional Pontiac Grand Prix which drove with the Porsche. We took the chance to compare the vehicles against each other considering all three cars drove together along the same route.

Toyota Prius ran the race at very respectable 52.9 mpg, while the Pontiac Grand Prix did 22.8 mpg, fairly typical for a midsize sedan. We observed that fuel economy measured in mpg is not as telling as fuel consumption, usually measured in gallons per 100 miles or in liters per 100 kilometers, a European metric. You can readily see a difference between equivalent fuel consumption of an electric, a hybrid and a conventional car on this plot.

More updates are coming about our recharging experience and about our greenhouse gas footprint.

Update: There was quite a bit of press coverage for this event.
One Gallon Challenge: alt-fuel "race" cars get up to 164 MPGe

Friday, August 21, 2009

The feasibility of rapidly charging EV's

Below is a poster I put together describing the feasibility of rapidly charging EVs.

There are three main things to consider: (1) Cell chemistry (2) Power source (3) Chargers and (4) Battery pack design.

The good news is that it's all totally feasible - though research & development is obviously required! A123 has already developed the cells necessary.

Remember that drivers would not rapid recharge all of the time. Most of the time they would charge EVs for the lowest cost - slowly - over night in their garage. But in the event they need the extra range, rapid charging should be an option.

Included is a chart Shane and I (with help from professional electrical engineer Dave Rodgers) put together. It shows how you could rapid recharge a motorcycle at home, while a sedan would require higher power - which is already available in most industrial buildings (think Home Depot, Best Buy, etc.).

Thus our vision is that the rapid recharge stations would not be at home - but rather at fueling stations like Hope Depot, Shell Gas Stations, etc. This is totally feasible - and the power already exists in many of these buildings.

(click to enlarge)

Electric motorcycle update

A couple years ago I built this all-new electric motorcycle - for $3,000 - using parts available on the internet:
Here she is parked outside the Caltech student machine shop - it's place of birth:

Now I'm back at MIT - and found some like minded enginerds that are interested in some e-motorcycle research.

Paul, Mike, Shane and I have been working on a rapid recharge pack for the above motorcycle (aka eMoto). We're designing/building a charger that will charge the pack in 10 minutes. The same design can be scaled up for any pack size - including a car.

Here are some pics of our very recent work:

Charger thoughts:

Battery welding (thanks to our sponsor Miyachi):

We're very constrained with space, etc. Here's a finished module (BMS boards sitting nearby - provided kindly by Texas Instruments - thank you!!):

One of the lithium modules on the motorcycle - datalogger on tank:

One module (as shown) - lead vs. lithium - you should feel the difference!
Lead acid: 333 Watt-hrs / 44 lbs of battery (actual)
A123 Lithium: 380 Watt-hrs / 10 lbs of battery (predicted - includes weight of enclosure, etc.)

Working on the charger (gray box w/ lid removed):

We're also using a commercial charger (which can't charge quite as quickly - roughly 20 mins) - with help from Manzanita Micro - (thanks Rich!).

Thursday, August 20, 2009

Quick Fabrication Update

We've got the new high voltage wiring routed in the car, the battery mounting enclosure is almost completed and Mike is tweaking the chain drive unit. There's no design work left at this point: the path to the finish line is clear, with a rush to finish the final assembly.

Sprockets and chain guard are in place, with Mike grinding a small channel to make room for the motor encoder wire.

Kevin, finishing the enclosure for the trunk-mounted battery pack.

The orange high-voltage (356 Volts) shielded battery cable runs from the rear to the front of the car, from the battery pack to the motor controller.

Arya has been working on getting the hood latch assembly back in place - it had to get cut off to make room for the motor controller enclosure (the clear plastic box to the right of the engine bay with the orange cables running into it).

Saturday, August 15, 2009

Porsche: 90.28 Miles

The Porsche has set its first calibrated range record: 90.28 miles!

The Porsche crew has been hard at work over the last month preparing to climb Mt. Washington and compete in the 1 Gallon Challenge.

Tyler, Mike, and I decided that Thursday was not the best day to do the hill climb, many of the concerns regarding braking and regen performance had not been adequately answered, and with only 3 people going, it didn't make sense to spend so much time driving there and back.

With a trailer and tow vehicle already secured, we decided to seize the opportunity to take the Porsche out on the 1 Gallon Challenge route and see how far we could safely go. We followed the route backwards starting from Cambridge and driving out there. We got stuck in traffic and took a long detour on the wrong Beacon St., stopped in a gas station to dodge a summer sprinkle (without a rain cloud in sight), the trailer got rear-ended and we spent a few hours eating ice cream while the police sorted out the paperwork. It was getting dark and cold, so we trailered the Porsche back to city lights and finished driving around town.

Heating only becomes an issue when driving in 3rd (i.e. high torque, low revs). We have motor temperature logs & need to analyze exactly the parameters when the motor overheats. It drives very well in 2nd, which is perfect for the country roads that are along the route. At the end of the drive, some capacity mismatch seemed to exist between the cell groups in the pack, but they all charged up perfectly balanced (so there isn't too much cell efficiency mismatch). Since this is the first time the pack has actually been balanced and fully cycled, we're hoping that some of the mismatch will work itself out as we cycle the pack more.

A lot of work has been going on in preparation for these drives, I'll detail more of what we've done in a coming post.

Saturday, August 8, 2009

Porsche Center Console Construction

Not to be outdone by the factory-built touchscreen console in the Milan, the Porsche crew built and installed a new center console for its touchscreen connected to the NI cRIO. The console was built by EVT's newest member, Amanda Turk, and team advisor Dan Lauber over a very productive weekend filled with tools and materials uncommon to the EVT.

We built the center console completely out of wood and woodworking bits, except for a few custom sheetmetal brackets. These are the same materials which the stock Porsche center console, on which our design was based, was made from. The center console is a highly visible and functional fixture in the car, so there was a laundry list of requirements in designing and building it. It had to look stock, not interfere with the operation of the accelerator, shifter, or adjustable seats, fit the National Instruments touch screen and cRIO controller, be sturdy enough for abuse, provide room inside for wiring and outside for the eventual addition of switches and indicators, and provide easy tool-less access to all the wiring and electronics inside it.

We mocked up the entire console in cardboard before making a run to Home Depot for lumber. The cRIO was mounted against the front wall, above the passenger's toes, we added a floor board under the passenger's feet to protect the wiring (it was originally just a big chunk of foam), and we re-routed a good number of the wires and cables.

One of the more interesting bits of construction, the touch panel was meant to mount to a much thinner panel, requiring the routing of the back side of the plywood panel with a 1/4" end mill. A couple layers of paint, and the center console was ready to show off at the open house and looks fantastic. Still to do is organizing all the wiring inside, and thanks to Amanda's mad sewing skills, a genuine viynl cover instead of paint. Also part of driver information/fit'n'finish is getting all the gauges working and looking stock.

Friday, August 7, 2009

Fabrication Progress

The motor/differential/speed reducer assembly and the motor controller (the two largest components in the engine bay) are mounted in their final locations. It barely fits - had the controller been 1 inch larger on any side and it wouldn't have fit in the engine bay unless the motor and transaxle were a parallel unit (motor laying flat, transversely in the car), with the controller laying flat on top of it. An advantage of the current configuration is that the motor, controller, gear drive and differential are all visible from the top, useful for outreach/demonstration.

Motor/speed reducer/differential assembly Version 2.0 is hanging from the lift; on the right you can see the motor controller (the black box) mounted in a polycarbonate enclosure (we went a little overkill on the waterproofing).

This photo is taken from inside of the engine compartment; you can see the clearance between the motor controller box and the top of the hood (the underside of the sheet metal) is about half of an inch.

The new motor frame, as seen above, clamped to the welding table - this is necessary to prevent the welding heat form warping the frame. This is version 2.0, which is slimmer and allows for the clearance to mount the motor controller next to it in parallel.

The components for the chain-drive speed reducer. From the left: flexplate adapter (connects the lower chain sprocket shaft to the differential); spline shaft and sprocket spacers (connects to the motor at the top); lower chain sprocket shaft with the lower bearing mounted and the other side of the triangular flex plate adapter.

Saturday, August 1, 2009

Nonstop Fabrication

We've been working around the clock (well, almost - a typical day starts at 10am and goes until 2am, including weekends) to get the fabrication done. We finally have the components we need and our CAD models are worked out, with the final assembly and mounting left. The pictures below may explain the lack of blog posts in the past few days.

The choice between spending more time in front of a computer vs. spending time in the shop with a TIG welder usually favors the latter: we can't wait to get this car on the road.

Mike, welding the coolant pump mount to the frame rail.

Matt, machining the 1/2 inch thick aluminum walls for the chain drive enclosure. The bearing plates for this cannot be allowed to flex at all, otherwise the chains (we're using 4 in parallel) will become unevenly loaded, leading to a cascading failure after the first chain snaps.

Arya, about to mount the wiring harness/interconnect box.

The motor / diff frame assembled and mounted in the car. A 1.7:1 ratio chain drive connects the two. This is our first iteration of this frame; currently we're working on the second one, that will be slimmer at the bottom and slightly offset to the left, to make room for the motor controller box.

A (very rough) mounting of the motor controller box. This will house the 640A, 400V controller electronics, and needs to be watertight. It's constructed from 1/4" polycarbonate, with a 3/8" sheet on the mounting surface of the controller. 1/2" thick polycarb is bulletproof.. It's waterjetted and dovetailed at the ends, which will be bonded and sealed with silicone.

Jigging up the motor/controller frame, version 2.0. This time we'll get it right . . . no warped joints, no flex, slimmer by 4.5" at the bottom, leaving more room in the engine bay for mounting other components and (future) battery modules.

The sponsor logos on the car, prepped for the open house (the shop was much cleaner for that event). We had the president, provost, supporting faculty and corporate sponsors stop by and attendance remained strong throughout the evening.

The custom 17-spline shaft for the motor is finally here. It was much cheaper to send the shaft out to be made via EDM (electrical discharge machining) rather than a conventional broaching method.

. . . more to come in the next few days