Case study: The Ecotek CB-26P

Case study: The Ecotek CB-26P

In general, I have avoided discussion of specific devices on these pages, and concentrated on generic technologies. But the case of the Ecotek CB-26P is especially interesting, since it has a very high profile on the Web and much background information is available, so this page focusses on this particular device.
Important background information. For the avoidance of doubt, readers should bear the following in mind:

1. This Case Study is not an unbiased review of the CB-26P; as is clear from other pages on this site, I have a very sceptical view of all aftermarket fuel "saving" devices, and this Case Study focusses on the negatives (ie what it does not do, rather than what it does do).

2. I have mainly considered issues of fuel economy (as the site name suggests), rather than the aspect of performance (in particular "driveability") that now forms the main selling point of the device.

3. Criticism of Ecotek's claims for the device is more aimed at their previous claims (which are unfortunately still widely repeated on the Net) rather than the current advertising.

Ecotek's CB-26P is an example of a device which bleeds air into the inlet manifold. For full details on this sort of device see this page, but in summary, it adds extra air to the engine under certain conditions and so adjusts the air:fuel ratio. On an engine that often runs "rich" (as many engines did before 1993), this could perhaps give genuine benefits.
It is instructive at this stage to consult Ecotek's patent application (pdf format). The important paragraphs are:

Typically the engine is tuned when at an idling speed, known as tickover, and the engine is tuned such that is operating optimally when ticking over. It is the only practical method of tuning the engine but once the engine is operated at speed and under varying load conditions it is not optimally tuned for such conditions and thus a degree of unburnt fuel is emitted from the exhaust of the car thus reducing the engine efficiency. 

There is provided a combustion mixture control apparatus for a spark ignition internal combustion engine comprising an auxilliary air valve ... [to] cause the air to fuel ratio taken in by the engine to be in the range of 13:1 to 15:1 and preferably in the range of 14.4:1 to 14.7:1.
The auxilliary air valve of the invention will act as a combustion mixture control apparatus to optimise or near optimise the air-fuel mixture in the manifold and thus the cylinders at all times during the operation of the engine. This in turn will reduce the emissions of unburnt fuel and partially combusted fuel from the exhaust of the engine thus reducing the emissions of carbon monoxide and hydrocarbons into the atmosphere. 

...and another half a dozen or so references to optimising the air/fuel ratio.
This wording in the patent application strongly suggests that the design purpose of the valve is, as I describe on my page, to lean the mixture off under certain conditions. There is some sense in this idea: some readers may remember having to turn the "mixture adjusting screw" at idle to get the right CO reading (which is a measure of air-fuel ratio or lambda). The lambda or air-fuel ratio achieved at loads and speed other than idle would depend on the carburettor jet / needle settings, and indeed frequently was not ideal.

Tellingly, until very recently the primary evidence for improved fuel consumption was the 1993 Warren Spring test data. This testing was carried out on a 1988 vehicle with a carburettor - this sort of vehicle might very well run too rich under normal operating conditions and so show some benefit when the Ecotek was added. The device was developed at a time when carburettors were almost universal.

So, that is all quite clear and consistent. But the critical point is that this absolutely does not describe how a modern (1993 onwards) engine works. The carburettor has been consigned to the dustbin, replaced by a computer-controlled injection system that allows the engine designer to precisely specify the desired lambda at all load and speed points. The lambda sensor in the exhaust corrects for any small deviations under most operating conditions, and indeed the "idle mixture adjustment control" simply does not exist on any post-1993 European petrol car. The mixture is already at optimum (or very nearly so) under almost all conditions, not just at idle, and so mixture adjustment as a way of significantly improving fuel economy is all but impossible.

Yet the CB-26P has not generally been marketed in the past as a device only suitable for pre-1993 vehicles, indeed Ecotek's technical pages have frequently actively denied that the device works by adjusting the air/fuel ratio. Instead the usual claim has been that the main effect is increased turbulence [swirl] (as you can see from the reviews) - which is not mentioned anywhere in this or any other patent application. Nor have Ecotek presented any test data to actually show that turbulence is increased (despite repeated challenges), although measurement of turbulence is a relatively straightforward test routinely used by many engine designers. A cynic might believe that is because they do not have any test data showing increased turbulence, which then makes you wonder how they can make such a claim.

It is true that adding in-cylinder turbulence can have a very slight benefit on fuel consumption and emissions - this page gives more information. But the maximum benefit is only around 2% given the reasonably high turbulence levels that exist in modern engines anyway. More fundamentally, the biggest effect of increased in-cylinder turbulence is a faster burn. By the time you have enough turbulence to affect economy, you must retard the ignition by typically 10 - 20 degrees to compensate for the increased burn rate.

A further claim typically made by Ecotek is that it improves fuel atomisation which makes the burn more complete. Again Ecotek do not present any test data to actually show that atomisation is improved, though this is not too difficult to directly measure. And it ignores the fact that atomisation on any reasonably modern engine is already very good, and the burn is typically 98% complete under normal cruising conditions. Thus the potential for fuel saving is extremely small. See this page for more information about fuel atomisation.
So, in summary:

• There are strong theoretical reasons to believe the CB-26P does not significantly increase turbulence
• There is no test data to show that it does increase turbulence
• Both theoretical and experimental understanding of engines says that increasing turbulence would not, in any case, give the 10-15% economy benefit typically claimed
• The device undoubtedly does lean off the mixture on most pre-1993 vehicles, and this could be responsible for at least part of the improvements observed

Turning away from the technical aspects, Ecotek set great store by the many magazine reviews shown on their web pages. Quite apart from the obvious point that negative reviews are not reproduced here, none of these reviews contain any scientific test data. We have emissions measurements at idle or fast idle, almost always without NOx, and fuel consumption measured on the road under totally uncontrolled conditions. And of course many of the positive comments, especially from the "modding" end of the market such as Max Power, are more to do with driving "feel" than economy or emissions. That the CB-26P can affect the "feel" of at least some cars is not in doubt, though the change may well not be for the better. Finally, many of the reviews and testimonials are for pre-1993 vehicles, and so may not be applicable to newer cars.
There is a general issue as well, that magazine writers are not normally technical engine experts and so may not be qualified to judge this or other fuel "saving" devices accurately. That's not meant as a personal criticism - nobody can be an expert on all aspects of cars, for example I know very little about ride & handling. But beware of assuming that all car magazine writers are technical experts on all aspects of vehicle engineering, as this is simply not the case.

A common opinion voiced by Ecotek as well as other makers of fuel "saving" devices is that everyone who tries them is convinced, while the critics are all people who have not tried them. And of course the makers and sellers of Ecotek and other similar devices display entirely positive testimonials on their sites, which seem persuasive at first glance. But the important thing to remember is that the reported benefits are almost entirely based on perceptions and uncontrolled tests, and hence show large variability (see this page for general comments on fuel consumption variability). It is not suprising that some users perceive a benefit, but equally it is not hard to find negative reports from actual users on the Internet. For example, from the Clio Sport forum:
Just removed the Ecotek valve. The car pulls harder, brakes better and runs quieter than before. If anyone has an Ecotek I'd strongly suggest removing it...let the ECU do what it's designed to do
Took mine off after running it for about 6 months, made a massive improvement. Waste of money!
People frequently ask why I criticise the CB-26P without having personally tested it. The reason is that I could fit one to my car but it would prove absolutely nothing. Any fuel economy effect would be lost in the variation of normal driving. The emissions tests I could do (basically CO and HC at idle, as measured on the MoT test) would say nothing at all about the effect on overall emissions while driving, and hence environmental impact. Proper performance testing would require multiple rolling-road tests with and without the device, which would be expensive and time-consuming. For these reasons, I also believe the customer testimonials often quoted by Ecotek to be of little value.

Readers looking for an independent view of the CB-26P may like to note that the Advertising Standards Authority ruled against Ecotek on 21st April 2004. The full adjudication is quite long, but in summary, the ASA believes Ecotek do not have any credible evidence that the device does in fact improve economy, emissions or performance. As a result Ecotek are effectively barred from making these claims in normal print advertising (magazines etc).

Important update - March 2005. For well over a year, this site has consistently stated that the CB-26P can have very little effect on the emissions or economy of lambda-controlled engines (post-1992 in Europe), for the reasons explained above. Ecotek have in the past always denied this, and as recently as December 2003 claimed fuel economy improvements of around 15% on "anything other than the very newest vehicles". They also "guaranteed" fuel economy and emissions benefits, and indeed many resellers and marketers typically claimed 15% economy benefits on all vehicles.
Under pressure from both the ASA and this site, Ecotek finally conducted a series of proper scientific emissions and economy measurements over standard drive cycles on modern cars. The results can now be found on Ecotek's site - and indeed prove that the CB-26P gives little or no economy and emissions benefit on a more modern vehicle.
The summary of the results is as follows:

• Fuel consumption (and hence of course CO2) around 1% better
• NOx generally slightly increased (about 5%)
• HC and CO generally slightly reduced (about 10%)

The critical result here is of course the improvement in fuel consumption of just one percent - something like a tenth of the benefit previously claimed by Ecotek. Even if this result represents a genuine technical improvement rather than simply random test-to-test variation (see comments on statistical significance), that means the device is extremely unlikely to ever pay for itself in fuel savings. Ecotek do claim that there will still be benefits on lambda-controlled cars "at higher speeds where the Lambda goes open-loop and the AFR controls switch off", but they do not seem to have any evidence for this. Lambda control is generally active under all cruising conditions, with the engine only going "open-loop" at close to full load. Under these high load conditions, there is almost no manifold vacuum to suck air through the CB-26P and so any effect on the engine will be minimal. Furthermore, it should be remembered that the EUDC test cycle over which these vehicles were run covers the speed range up to 120 km/h (75 mph) and so cannot be considered as just low load and speed. (This is especially true given that the cars tested were relatively low-powered; the EUDC cycle demands acceleration from 62 to 75 mph in 20 seconds, while the Ford tested takes over 10 seconds to do this absolutely flat-out.) If the device really could work in the real world, one would expect to see the benefit on this test as well.
An additional argument made by Ecotek is that the device will give benefits on 1993-1995 "Euro 1" engines, since the lambda control is in some way less effective than on the "Euro 3" cars tested, hence allowing the CB-26P to "manipulate" the AFR more. Having worked on engines covering Euro 1 to Euro 4, this seems highly unlikely - the fundamentals of lambda control are the same on all catalysed vehicles, and any differences are likely to be small. One might also wonder why Ecotek did not simply choose to include a "Euro 1" vehicle in the test group, if this really is such an important factor.
In terms of toxic emissions (HC, CO, NOx), there is no clear trend - the best that can be said is that emissions with and without the device fitted are broadly similar. An obvious conclusion from this is that the CB-26P is not significantly affecting the combustion process (either for better or for worse), which again suggests that Ecotek's claims of swirl and turbulence effects are untrue. The lack of a significant rise in NOx shows that the lambda controller in the ECU is doing its job and compensating for the extra air flowing through the device.

In summary: even on cars chosen by Ecotek, with the CB-26P tuned by them and the tests under their overall control, they can only demonstrate tiny economy and emissions benefits on a lambda-controlled vehicle. The sceptics, such as myself and the ASA, have once again been proved correct!
A more general point is the relevance to other fuel "saving" devices. If you search the Internet you will find hundreds of people convinced that they have seen 10%, 20% or even greater fuel economy benefits on lambda-controlled cars through fitting the CB-26P. Yet when the device is scientifically tested - all conditions kept constant, except for the presence or absence of the device - the apparent benefits all but disappear. The conclusion must be that users who reported these large economy improvements were either mistaken, or were fooled by other changes such as driving style, journey type, etc. This should be borne in mind when looking at "testimonial" evidence for other fuel "saving" products.
To be fair to Ecotek, since this test data was obtained they have shifted the advertising focus of the device to its performance benefits, and have largely dropped claims that it will save useful amounts of petrol (or greatly reduce emissions) on lambda-controlled vehicles. Unfortunately not all marketers and resellers of the product are so scrupulous, and it is still common to find "saves 15% on ANY car!" as a claim for the device. Such claims should clearly be taken with a very large pinch of salt.

The other aspect of the Ecotek CB-26P is the claims of increased performance. Most fuel "saving" devices make some claim of a performance increase, but the CB-26P is regarded by many as a primarily a performance mod, with the fuel saving almost a side benefit.
The argument why it can have very little (if any) effect on performance is set out in detail on this page; in summary, there is very little manifold vacuum at wide-open throttle, and so the air flow through the device is negligible and its effects essentially zero. (As an aside, worries that the device will lean off the mixture at full throttle and cause overheating damage are therefore probably groundless.) In fact, even Ecotek themselves do not actively claim a performance increase in the usual sense - that is to say, an increase in engine power or torque. The only rolling-road data on their website (from VW Motoring) quotes a 1 to 2% increase in BHP, which would be totally imperceptible to the average driver and almost certainly within the test-to-test variability of the rolling road.
Why, then, do so many users report that their car feels faster? A large part may well be psychological - especially since the device makes a loud "slurping" noise while working. Modifications, particularly ones that change the volume or character of the engine's sound, are widely considered to lead to an improvement in perceived performance. However the device does have an effect on the "driveability" of the car, which may be perceived as a performance boost.
On many modern cars, there is no direct connection between the accelerator pedal and the throttle blade. The pedal sends a signal to the ECU, which moves the blade as it sees fit (the so-called "fly-by-wire" system). This allows the engine designer free choice as to how the blade should react to particular pedal inputs, and this "driveability calibration" is responsible for much of the "feel" of the car. For example, when the driver suddenly lifts off the pedal (to slow down or change gear), there are two possible responses:
1. Snap the throttle blade shut immediately, bringing the engine quickly down to its minimum torque output. This gives a very "sharp" response and plenty of engine braking, which is often desired. But because the engine has been so strongly throttled, there is a slight delay after the driver re-presses the pedal before full torque is available again. This is sometimes referred to as a "flat spot", and is particularly evident on less sophisticated systems that struggle to keep emissions under control during rapid throttle movements.
2. Close the blade gradually, and not so fully (on non-fly-by-wire systems, a similar trick can be pulled with the idle control valve). This keeps a higher engine torque and so the response to the pedal being pressed again is more rapid. However, it gives a feeling that the car is "running on" with no engine braking, which many drivers find very unpleasant, and also slightly increases fuel consumption.
The ideal response is entirely down to the preference of the driver. It is however true that behaviour (2) gives smoother power delivery through gearchanges (since the engine doesn't "die" during the change) and so may give a slight improvement in through-the-gears acceleration. Engine designers look for a compromise between (1) and (2) to give sufficient engine braking and smooth gear changes, but do not always succeed.
How is this relevant to Ecotek? Basically, the CB-26P valve acts as a leak round the throttle, simulating a larger throttle opening and so changing from behaviour (1) to behaviour (2). If your car is set towards (1), and if you prefer (2), then you may like how your car drives with the Ecotek better. But this not a "real" performance increase in the "traditional" sense, as the rolling-road data mentioned above makes clear.

Important update - March 2005. For well over a year, this site has consistently stated that the CB-26P can have very little effect on "real" performance, for the reasons described above - the only likely benefit would be slightly faster through-the-gears acceleration due to less "throttling" during the gearchange. In Feburary 2005 Ecotek carried out a series of performance tests, the results of which can be found here. In many ways this test program was very well designed and carried out - the Datron equipment used is highly regarded in the automotive industry, and by evaluating the results from 10 mph rather than rest the inevitably highly variable "take-off" is eliminated.
Unfortunately one critical aspect of the testing was missed - the A-B-A check. It is not sufficient to show an improvement when a device is fitted; you must then remove it again and show that the improvement goes away. Otherwise it is possible that some entirely unrelated external factor was responsible for the change. In the case of Ecotek's performance testing, this factor might be:

• Wind (even just 10 mph would have a very significant effect at higher speeds)
• Temperature of oil, coolant, tyres, etc
• Fuel level (just one gallon of petrol - easily used during the course of performance testing like this - would alter the weight and hence acceleration of these cars by about 0.5%)
• etc

Of course, quite possibly none of these factors changed during the testing - but without an A-B-A test it is impossible to know. Even taking the results entirely at face value, it is clear that the absolute improvement in performance is not highly significant - about 3% in most cases. To put that into context, it would equate to:

• Adding two additional horsepower to the engine
• Lowering air temperature by 10 celcius
• Raising air pressure by 25 mbar
• Removing about 5 gallons of petrol from the tank

Most drivers would find it impossible to detect a change of this magnitude; certainly the test results do not appear to justify Ecotek's previous claims of "transformed" performance. The implication is that the improved performance noted by some Ecotek users is much more due to altered "feel" and speed of throttle response, rather than actual objectively better performance. Some drivers undoubtedly do prefer this altered feel, and for them the CB-26P probably does justify its purchase price. But it should always be understood that this is primarily a change in the character rather than amount of performance. (More positively, this change in performance characteristic does not seem to come at the expense of increased fuel consumption or emissions - unlike some performance modifications).

Footnote
Ecotek's lawyers wrote to me at the end of April 2005 to complain that this site was "unbalanced" in its treatment of the CB-26P. It is of course true that I have focussed here on the negative aspects, though I would strongly deny that any of the analysis is incorrect or different to normal industrial and scientific practice. However, in the interests of fairness, I suggest that potential customers either visit Ecotek's own web page, or indeed their Forum, for the "other side" of the story. Suggestions from Ecotek for changes to this site to improve accuracy remain welcome, although I make no guarantee that they will be implemented.

Please also read the general comments on fuel "saving" devices, if you have not done so already