Why Water Won't Improve Your MPG

I get mail. Hundreds of pieces of mail every month, and that includes e-mail, paper mail, the occasional voice mail and even a smattering of faxes. For the last two years, an unhealthy proportion of that correspondence has been about the same thing—making your car run on water instead of gasoline. With fuel escalating to historically high prices, followed by the global collapse of virtually every market, people are inclined to search for a way to reduce their monthly expenditures for gasoline or diesel. Which, of course, is perfectly understandable.
Over the years, I've tested plenty of gadgets that purport to reduce fuel consumption. None of them worked. None. Lately, I've tinkered with a number of them that rely on the same principle: using electricity from the car's battery to electrolyze water in an onboard cell and burning the resultant hydrogen-oxygen mix in the engine. In theory, the burning hydrogen will provide extra energy, reducing the amount of gasoline you need to move on down the road. There are dozens of websites, and dozens of people on Ebay touting these devices, guaranteeing, depending on their level of chutzpah, anywhere from 15 percent to 300 percent improvement in fuel economy by simply bolting on one of their devices.
This doesn't exactly rival string theory in it's complexity. You can build a serviceable hydrogen generator from an old peanut-butter jar and some leftover copper pipe or roof flashing. There are plans to construct this device that you can get online, too. Just use some aquarium tubing to duct the hydrogen-oxygen mix (usually abbreviated as HHO) into the intake manifold, and you'll see the gas gauge stay at "Full" a bit longer—or so they say.
When these devices first hit the Net, I had an immediate opinion: Rubbish. I discussed the theoretical science a while ago. It's bad science. This malarkey boiled down to perpetual motion: something for nothing. Essentially, it takes more energy—in the form of the chemical energy in the gasoline you're burning in the engine, to spin the alternator to make the electricity and generate the HHO—than you get back. In fact, it's not even close: Multiply all the inefficiencies in that system and you only get a few percent back, certainly not in excess of 100 percent.
Two things happened after I said that. One, I got overwhelmed with mail from true believers who volunteered to have me test their car. And a lot more from people who accused me of being in the employ of the auto and petroleum companies, suppressing this breakthrough technology and keeping the American public enslaved. Sigh. Were this true, I'd be living in a much nicer house.
The second thing I noticed in the last 6 months was a change in the claims made by HHO proponents. The extra fuel economy was supposed to come not from the additional energy contained in the hydrogen, but from the hydrogen's ability to facilitate the combustion process, producing more power from the engine with the same amount of gasoline. Which is also malarkey. Before you start e-mailing me copies of those same scientific papers (I've seen them a dozen times) that supposedly prove that this works, let me tell you, these documents don't apply to your car. Without getting very detailed, these papers all deal with ultralean experimental engines with fuel-delivery systems enhanced with a stream of pure hydrogen, achieving a small improvement. They have nothing to do with retrofitting a conventional engine (with computer-controlled engine management that keeps the mixture near a perfect 14.2:1) with a device that adds a hydrogen-oxygen mix.
One point of interest: A conventional car engine ingests on the high side of 500 liters of air per minute at idle, and a great deal more at highway speed. These generators generally produce a liter or less of HHO per minute. Or roughly 50 liters per hour, of which only two-thirds is hydrogen. At atmospheric pressure, hydrogen has a density of 0.0899 g/liter. One NASA study used 640 grams of hydrogen per hour to sweeten the mixture for its conclusions. I'll leave the homework to you, but, basically, the amount of hydrogen added to the combustion process by onboard hydrogen generators is far smaller than one percent of that used by the studies that hydrogen-enrichment proponents are quoting as "proof" that their gadgets work. Could you make a hydrogen generator that made that much HHO? Sure, but it would be huge, use far more electricity than the onboard generator could possibly produce, and consume most of the power the engine put out—and it would still not improve fuel economy
I've been tinkering with a couple of homemade and commercial HHO generators. I have instrumented several cars with HHO generators I can switch on and off, flow meters, scan tools and instantaneous mileage displays. I've tested them on the road and on chassis dynamometers, and have never seen any improvement. None.
Of course, when I finally got tired of answering letters from HHO proponents and stopped posting results on our website, the buzz on the internet took a predictable turn: Commenters and letter-writers claimed that I had discovered that the things actually worked, and had stopped writing because I was embarrassed to admit my error.
More rubbish.
Here's what's really been happening. I've been working with NBC's Dateline to debunk the whole hydrogen-on-demand industry. The show's producer bought a car, an ordinary five-year old Honda Accord, to perform our tests. I checked the car over to make sure it was up to spec. Then we did some over-the-road and steady-state dynamometer testing to establish base-line fuel economy numbers. NBC followed my testing up with additional testing at an EPA-certified emissions lab, which wasn't cheap. The lab used its climate-controlled emissions dyno to establish fuel economy numbers in our Accord with the same protocols the EPA uses to generate the numbers on the window sticker of new cars. They're accurate and reproducible to well under 1 percent.
Then we took the car to a specialist who installed, for nearly $1900(!), a hydrogen generator and a system of other enhancements. There was a fuel heater, fuel-line magnets (which I debunked here), and several inscrutable boxes full of electronics designed to fool the car's computer into using less fuel. There was even a bottle of acetone to add to the fuel. (This is something that I've mentioned doesn't work here and here). The specialist guaranteed major improvements in fuel consumption. One week and nearly two grand later, the producer from NBC (who still hadn't identified himself as anyone except a guy who was tired of spending $50 to fill up his tank) picked up the car. He got a gas receipt proving the installer had seen 96 mpg, nearly triple the original economy.
We took the car straight back to that same EPA lab for another round of testing. It was followed shortly by a week's worth of road testing, dyno testing and general poking about to see what we could discover.
You can guess, right? The total improvement in fuel economy after $1800 plus of expenditure? Bupkis. Too small to measure. Nada. In fact, if you look at the EPA tests with the system switched on and then off, there's a tiny increase in fuel consumption when the system is turned on. I attribute this to the 15 amps or so of current the electrolysis cell consumes to produce hydrogen. That current uses horsepower to spin the generator, and that consumes gasoline. The hydrogen "boost" couldn't even compensate for its own losses.
And that is exactly what I've been saying for years. These systems don't work.

How to Monitor Your Fuel Economy in Real Time on the Road

With gas hitting a national average of $4 a gallon, everyone's been talking about how to improve fuel economy—myself included, be it for taking summer road trips or avoiding goofy (and expensive) gas-saving gadgets. But how do we know if the fuel economy we're getting is accurate, whether we're driving an unmodified car—or one equipped with a supposed miracle product?
On the road, most drivers just read the trip odometer and divide by the amount of gas they purchase. Simple enough, eh? Sure is, but that doesn't account for differences in the driving cycle during each individual tankful. And that can vary an enormous amount. Of course, over a few tankfuls, the variances will average out, right? Not exactly. Your vehicle will get poorer fuel economy during the winter. That's because of the increased electrical use for lights, wipers, heat and longer warmups—not to mention the extra drag caused by moving snow and slush out of the way of the tires. Spring and fall are good, but A/C use can certainly cause a mileage hit during the hot summer months.
In the back-and-forth my weekly repair Q&A, one reader described to me his supposedly foolproof method for checking mileage—and subsequently the accuracy of his gas-saver gadget. It goes like this: He drove a delivery van on the exact same route twice a day. So he topped off the tank one morning, and drove the route like he normally does. He then installed a pretty pricey magnet (over $100) on the fuel line during his lunch hour, topped off the tank and drove the route again that afternoon. His mileage increased 12 percent, thereby proving his gadget worked. Right?
Well, even this sharp reader failed to account for the fact that in the cool morning temperatures, his fuel tank was also cooler and denser. It took a certain amount of fuel to fill the tank from the day before. And as he drove his route, the fuel heated up and expanded, largely because fuel pumps work continuously. Even then, only the excess fuel (heated up as it passes through the engine compartment) is returned to the tank. Consequently, it took a smaller amount of fuel to fill his tank. So he made the afternoon run with a warmer tank ... which he didn't refill until the next morning when the tank was cool. That cooler fuel had shrunk, making it appear that he had used less.
At my suggestion, our dear reader repeated the cycle with the magnet installed only during the afternoon runs. His results, it turned out, were very different. I had a feeling they would be.
But there's a whole lot more going on here. I'd argue that the mere expectation of increased mileage will actually make your mileage increase. I've seen it happen a thousand times: If you know the fuel economy and a gas-saving gadget are under scrutiny, you'll accelerate smoother—and probably drive a bit slower, too. I call it “voodoo mileage.” To find accurate results, a driver must remain ignorant not only to how much fuel is being consumed, but also to the entire experiment. It would have to be a true double-blind study.
Here at PM, we report fuel economy numbers on our long-term fleet by averaging the economy over a period of some months, and from a number of our staff's driving experiences. When we do a comparison test, we'll drive all of the vehicles on the same day, at the same time, at the same pace, on a several-hundred-mile loop. We even swap drivers every half-hour to equalize driving styles.
There are some tech tricks you can use to help monitor your fuel economy as you drive. I'm fooling around with a couple of devices right now. I just installed a Scangauge on my motorcycle. This $180 device plugs into the On Board Diagnostic System (OBD II) port under the dash of virtually any post-1996 car or light truck. It operates as a scan tool (much more on that here), so it gives me trouble codes and streaming data, but it also works as an electronic gauge cluster and trip computer. I can track battery voltage, coolant temp and sundry while tooling down the interstate. But one feature is invaluable: the instantaneous fuel economy readout. It gets fuel quantity data from the injection timing—the longer the injectors are open, the more fuel they squirt. It's amazingly accurate, too.
And I've just started tinkering with another gadget with some similar features: the CAMP2 from HKS. It's a scan tool/gauge package/trip computer like the Scangauge, but it uses either the car's internal dashboard display or an aftermarket TV screen of any sort. It's intended to be professionally installed, but I've made a portable enclosure that I can suction-cup to the windscreen with RAM mounts. Unlike the Scangauge, the CAMP 2 has a graphic display that can be configured to reflect a dizzying number of parameters available from the vehicle's OBD II system. Input the car's weight, and you can even get an instantaneous horsepower indication. You can look at raw numbers, or at a simple analog-style gauge with a moving needle. But my favorite is the display that gives you something looking more like an oscilloscope trace, showing you what any given engine parameter is for the last few seconds or minutes—including rewind capability, so you can review after you've pulled over to a safe place.
If you'd like to alter your driving style to achieve high fuel economy, these devices are unparalleled. The slightest upgrade, downgrade or movement of your right foot leaves a dent or a bump in the economy trace. I've only begun to investigate its potential, so stay tuned.
Don't feel like dropping hundreds of dollars for one of these high-tech gadgets? Pick up an old-fashioned vacuum gauge at the local parts store. Monitoring manifold vacuum as you drive around will give you a fairly clear picture of your instantaneous fuel economy. BMW models have had a vacuum gauge integrated into the instrument panel for generations—it's simply labeled in miles per gallon instead of inches of vacuum. Higher manifold vacuum means higher mileage—the needle will sink alarmingly as you open the throttle, which will soon teach you to featherfoot. That's good advice, because altering your driving style is probably the least expensive and most effective gas saver of all.