A better way to think of bmep is work per unit displacement per engine cycle, that gives units of pressure, but explains why we use it, as it gives work (proportional to torque) per cycle, independent of engine size, thus one can compare engines of different sizes.

One can show that bmep = (vol eff)x(intake density)x(Fuel air ratio)x(Lower heating value in kJ/kg)x(indicated thermal efficiency)- fmep

For full load, as you apparently have, the fmep is only dependent on engine speed, and should be the same for all fuels.  The engine volumetric efficiency is a function of engine speed for a given cam, and should also be the same for all fuels, The indicated efficiency should close to the same for any fuel, unless you changed spark timing. So the main things different are the density, fuel air ratio and heating value.  Intake density would be affected somewhat by fuel evaporation (temperature effect, as alcohols have higher heats of vaporization), depending on what kind of fuel injection you use.  But the main things active here are the fuel air ratio and the heating value, the later of which should be easily calculated.  Your results should then obey the above relationship.  So you need to look at the fuel air ratio, which needs to be measured, and be sure that your data is reliable and repeatable.  If you take more data, be sure to monitor fuel air ratio, and intake manifold temperature and pressure.

BMEP is expressed as a constant pressure on the piston to give an engine power which is the equivalent of a real-cycle throughout a stroke. When alcohol fuels are used in especially naturally-aspirated engine, they are increased the engine's volumetric efficiency due to their properties such as latent heat of vaporization and boiling point. This situation has led to an increase in BMEP. We also need to know that the test conditions to interpret the graph.

A better way to think of bmep is work per unit displacement per engine cycle, that gives units of pressure, but explains why we use it, as it gives work (proportional to torque) per cycle, independent of engine size, thus one can compare engines of different sizes.

One can show that bmep = (vol eff)x(intake density)x(Fuel air ratio)x(Lower heating value in kJ/kg)x(indicated thermal efficiency)- fmep

For full load, as you apparently have, the fmep is only dependent on engine speed, and should be the same for all fuels.  The engine volumetric efficiency is a function of engine speed for a given cam, and should also be the same for all fuels, The indicated efficiency should close to the same for any fuel, unless you changed spark timing. So the main things different are the density, fuel air ratio and heating value.  Intake density would be affected somewhat by fuel evaporation (temperature effect, as alcohols have higher heats of vaporization), depending on what kind of fuel injection you use.  But the main things active here are the fuel air ratio and the heating value, the later of which should be easily calculated.  Your results should then obey the above relationship.  So you need to look at the fuel air ratio, which needs to be measured, and be sure that your data is reliable and repeatable.  If you take more data, be sure to monitor fuel air ratio, and intake manifold temperature and pressure.

Explanation of  prof. Sornsom  is sufficient to compare the obtained graphs, check all the parameters that determine the operating conditions of the engine.

Providing that the results are repeatable and the test conditions are the same, presented graphs can be read as follow:

Due to all the phenomenons described above in details by prof. Sorenson, the engine developped different torque ( that is proportional to BMEP) at specific revolutions for different types of fuels. Colloquially one can say that properties of the fuel E10 gives the engine chance to develop a higher torque than others, the specific fuel oil conssumption is lowest as well. It can be an advantage, providing that all other operational parameters of the engine are within the acceptable limits and/or parameters like exhaust gas emission are not compromised. In order to evaluate the test, many other aspects should be taken into account.

Engine performance is classified into three catogoriesfor constant speed engines

Measuring performance parameters, (BTE, EGT  and volumetric efficiency) varied with BMEP

Variation of pollution levels of particulate emissions, oxides of nitrogen and CO for diesel engine with BMEP

Variation of pollution levels of UBHC, CO and oxides of nitrogenwith B MEP for gasoline engine

Measuring combustion  characterristcis like PP, TOPP and MRPR, ignition delay

combustion duration at full load operation 

BMEP is defined as specific torque, toque per unit volume. It should be taken on X axis.  BMEP is important parameter for constant speed engines.   

All other parameters like BTE, VE, EGT, CO, UBHC, NOx should be taken on Y axis for constant speed engines.

When same engine is tested with different fuels with different properties, then BSEC, brake specific energy consumption defined as eergy consumed by the engine in producing unit brake power is to be evaluated (BSEC= 1/BTE) Lesser the BSEC, the better the performane of the engine. 

If different engines with different configurations are to be tested with same fuel, then BSFC is to be evaluated. 

For variable speed engines. speed is to taken on X axis and all other pearameters are to be taken on Y axis.  For Variable speed engines, speed is important parameter in evauating the performance of the engine. 

I strongly agree with the explanation of prof. Sornsom.

I just hope to add similar view point regarding setup. Generally, if the engine is allowed to run at specific speed to provide a specific power, there will be different BSFC, and power or BMEP will be the same with minor differences due to instrumentation errors and engine variations. However, if the test is carried out such that at specific engine speed the throttle opening or fuel pump is regulated, there will be different power delivered by the engine at fixed opening/fuel amount that will depend mainly on fuel type, and so the BMEP will also depend mainly on the fuel type. Both test cases are differe and so researcher should be care which test conditions he followed, correspondingly he will be able to discuss the received results.

Yes, Prof Sorenson answer is agreeable. For comparison of different fuels, we have set the load and speed rather than only constant speed and throttle, Coz throttles sometimes vary and not able to be kept at constant setting. However, if the load (Nm) and speed can be constant, the BMEP can be easily differentiated. As all the fuels should produce same BMEP, but we will see really they are producing same power but not same BMEP meaning pressure-acting on the piston not same fuel to fuel.