What is Corrected Horsepower?

We have all seen & made claims of an engine’s horsepower. However, this stated horsepower is almost never what the engine actually made for power. How can that be? Most of the stated horsepower numbers are “Corrected” values. The correction standards were developed to discount the observed horsepower readings taken at different locations & weather conditions. It is obvious that an engine builder on Mt Everest could not produce as much horsepower as a workshop in Sydney at sea level. There is just less oxygen for the engine to burn at the higher altitude. What are less obvious are the other weather condition effects on the engine. So in order to compensate for this all advertised horsepower is “corrected” to several different industry standards. 

Most of you know about Atmospheric Correction Factors that are used to compare an engines power output for one day or location to another. However, these factors can be rather confusing & even deceptive. Everybody seems to declare that their engine’s horsepower as the “etched in stone” number, however we also know that the engine will make very different power on different days. Excluding other factors like engine temperature & quality of fuel used, the engine output is dependent on the amount of oxygen in the air. So the only way to compare an engine’s horsepower is to correct the output on any given day to some standard.

The most common are the SAE standards. The older J607 standard considers that the engine was run on a 60°F day with 0% humidity & barometric pressure of 29.92 in-Hg or the newer SAE J1349 standard of 77°F (25°C) day with 0% humidity & a barometric pressure of 29.234 in-Hg (99 KPa). Also the ECE standard is the same as the SAE J1349, but does not use mechanical efficiency in the calculations. The DIN standard which corrects to 68°F (20° C) day with 0% humidity & a barometric pressure of 29.92 in-Hg (101.3 KPa) & the JIS standard corrects 77°F (25° C) day with 0% humidity & a barometric pressure of 29.234 in-Hg (99 KPa), but uses different correction curves than the others (as a substitution for using mechanical efficiency factors). Further, we have the J1995 corrects 77°F (25° C) day with 0% humidity & a barometric pressure of 29.53 in-Hg (100 KPa).

Since very few engines are actually run in these conditions we apply these correction factors so that it is possible to compare the results taken on different days. First let’s just look at the weather correction, we will see the second section dealing with mechanical efficiency later. Consider if you take a baseline run of a normally aspirated four stroke V-8 engine on a sultry day in late August, say 85°F & 85% humidity & 28.85 in-Hg & the engine produced 400 Hp. Then after you finished making all your modifications you retest the engine in late September when it is 55°F & 35% humidity & 30.10 in-Hg, the engine now makes 442 Hp. That’s almost an 11 percent increase in HP, however the engine is actually producing the exact same amount of horsepower according to the J607 correction values of 400 HP * 1.1005 ≈ 440 HP & 442 * 0.994 ≈ 440 Hp. If you had retested the engine in the same weather conditions it would have made 400 HP again. 

 

There are many different correction “Standards” out there, but here is the SAE J1349 formula:

cf is the final correction factor multiplier
Pd is the pressure of dry air in hPa
(990 hPA = 99 kPa)
Tc is the air's temperature in degrees Celsius

One more source of confusion about the SAE J1349 is all the different values quoted for the Barometric Pressure in inches of Mercury. If you search around you will find the base values are different. Some will quote 29.234 in-Hg & others 29.318 & others 29.380. How can they all be correct? Well the calculations are done in KPa or millibars. These units are all true pressures, however inches of mercury, although considered a pressure unit, changes with temperature. This is because mercury expands as it gets warmer. Therefore 99 KPa at 32°F is 29.234 in-Hg & 99 KPa at 60°F is 29.318 in-Hg.

Now this may sound confusing, but these formulas were developed to attempt to allow standardize advertised HP ratings & comparisons. The formulas are based on the amount of oxygen that is found in the air that the engine is breathing. The greater oxygen the more fuel can be burned & thus more horsepower. However, these formulas are not perfect. They were developed empirically & are a good approximation for the variables of humidity, temperature, & absolute pressure. However, internal combustion engines develop power on many other variables & although it is possible to have the same correction factor at high temperature & pressure as low temperature & pressure, the engine will make different power. The wetting effect & temperature differences are not perfectly compensated for. This gives rise to the “purist” touting that all engines must be tested at the same atmospheric conditions or else the results are useless. In a prefect world this would be true, but this would be ludicrous. The cost of building an environmentally standardized test cell is well beyond the capabilities & cost of even large OEM companies & would give rise to even more deception in horsepower advertising. 

Now lets consider the next effect on the SAE standard that some other industrial standards do not include, the “Mechanical Efficiency” of the engine. Which is basically the amount of energy the engine got from the fuel versus how much energy actually was produced at the flywheel. This is a measure that includes the frictional torque, viscous effect, etc. required to rotate the engine. If we take the SAE standard that a four stroke normally aspirated engine consumes 15% of its’ developed horsepower to rotate the engine.  This is another huge point of debate, but it does make sense. If we want to correct the observed horsepower to a standard condition, it make sense that the friction required to rotate the engine does not change with added oxygen in the air. So in the last example the engine produce 400 Hp on that hot August day. This time consider the SAE J1349 correction standard which has a correction factor of 1.0634. According to the SAE 15% standard it took 70.58 Hp (400 / 0.85 – 400 = 70.58) to overcome the friction from ring drag, bearings, valve train, etc. Since this is a constant value no matter where the dyno test was taken, we know that the energy produced by the engine was actually 400 + 70.58 = 470.58 Hp. Now if we want to compensate for the atmospheric condition then we should use the amount of energy that the engine got from the fuel supply. So we take the 470.58 Hp * 1.0634 = 500.42 & then subtract out the constant Hp reading of 70.58. 500.42 – 70.58 ≈ 430 Hp. 

Now it does make sense that the frictional torque is almost constant no matter how much oxygen was in the air, but the SAE flat rate 15% does not accurately cover all internal combustion engines. It is a compromise. In the example above we used a normally aspirated 4 stroke V-8 engine, but what if it were a two stroke V-8 outboard engine. It is quite obvious that the two stroke has much less frictional drag. It has no camshaft, timing chain, valves & springs, oil in the crankcase, etc. Comparing these two engines with the same 15% friction losses does not work. That is why some higher end dynoing software calculate the friction losses on many different variables, like the displacement, stroke for piston speed, type of aspiration, number of strokes, type of fuel, & RPM. Using this information will yield much greater accuracy in calculating a mechanical efficiency & therefore a much greater accuracy for in house comparisons between pulls. However, in order to advertise the value as SAE J1349 compliant you must usually use the SAE mechanical efficiency number. 

Another way to get accurate mechanical efficiency is to use a dyno that can “motor” the engine, like an AC dyno. Just measure the amount of power it takes to drive the engine & then use those values for your own custom mechanical efficiency. Once again though, you will need a high-end software package that will easily allow you to use the new efficiency or else you will be doing a lot of tedious & time-consuming hand calculations. But once again, this solution is not perfect either. Many will argue correctly that motoring the engine is not the same because there was no heat, bearing loads, metal deformation, etc. 

Some companies who are working on a particular engine family will actually test the same engine under many different conditions & develop their own correction table. To these companies it is vital to know how their engines will perform under specific varying conditions. Consider snowmobiles that will operate at many different altitudes & temperatures, but they can pretty much discount the effects of humidity because the engine will almost always operate at temperatures below freezing. However, it is critical that their engines perform well at extremely different barometric pressures. An exhaust designed to run at sea level will not perform well at all in the mountains. Further, the opposite is true for marine engines. These engines are run most often at sea level, warm temperatures, & high humidity. Or a waste gated turbo engine that is pretty much impervious to even large barometric pressure changes. Thus the one size fits all SAE approach does not work well. 

The debate over the validity of correction factors still lingers on, but they are the only way to make realistic comparison of your engines on different days. There are, & always will be, unscrupulous competitors who advertise inflated horsepower gains by manipulating the correction factors, however they are eventually exposed at the races where it counts to the customer. In order to perform accurate & credible results you must use some factors & try to conduct your tests under “similar” test conditions. In fact, SAE requires that the corrections be less than ± 7%. So in the example above we would not be allowed to use the STD or standard J607 SAE factor of 1.1005 because it is correcting by more than 10%, however the SAE J1349 factor of 1.0634 could just barely be used. 

Now that the importance of these correction factors is known they must be entered accurately for your test be to considered valid. Although the formulas look complicated you don’t really have to know them, because any dynoing software worth using will do it for you based on the three environmental variables of temperature, humidity, & absolute barometric pressure. Note that you must enter the absolute barometric pressure NOT the relative pressure based on altitude, this can also be a source of confusion. Unless you are at sea level the barometric pressure that the weatherman states has been altitude corrected & you must use the actual pressure. Once again, most dynoing software will be able to do the conversion for you. Also be sure to enter these values at the beginning of the test after the dyno cell has come up to a stable temperature. Failure to do this will show lower horsepower than your engine actually made. Once again you should consider finding a dyno that will automatically enter these values for you, because many times you will forget to write them down & that will invalidate the dyno pull that you just made & could even lead you to discounting a modification that did actually increase the power of you engine. Also, for advanced software that use more realistic mechanical efficiency you must enter the required information about the engine, such as bore, stroke, number of piston, type of engine, etc. 

It is also important that you use the same correction method for all testing & that your customer is shown the correction method used to calculate the horsepower. The customer may not understand all that went into the horsepower reading, but at least you will know that service was provided correctly & honestly. When considering a dyno you should research how the companies allow you to do your corrections. It may not be important now to be able to enter custom correct factor or even enter any at all, but it most likely will be later on down the road.

 

Note: The new SAE J2723 is actually not a new set of correction factors, it is simply a new procedure for using the existing factors (J1349 & J1995) used by automotive manufacturers. See the paragraph below from the SAE's website: Power & torque certification provide a means for a manufacturer to assure a customer that the engine they purchase delivers the advertised performance. This SAE Standard has been written to provide manufacturers with a method of certifying the power of engines to SAE J1349 or SAE J1995. Document SAE J2723 specifies the procedure to be used for a manufacturer to certify the net power & torque rating of a production engine according to SAE J1349 or the gross engine power of a production engine according to SAE J1995. Manufacturers who advertise their engine power & torque ratings as Certified to SAE J1349 or SAE J1995 shall follow this procedure. Certification of engine power & torque to SAE J1349 or SAE J1995 is voluntary, however, this power certification process is mandatory for those advertising power ratings as "Certified to SAE J1349".

ref: Modified Abstract from Land and Sea.