Battery Electric Vehicles in Heat Simulation

[Customer]

Hi, 

How do I deal with Battery Electric Vehicles (BEV's) in my Ventsim Heat Simulation? What are the key differences with how diesel vehicles are handled? 

[Craig Stewart]

Battery Electric Vehicle (BEV) heat should be based on the continuous power output (CPO) of the machine.  Peak power output is another figure sometimes quoted in specifications, but should be avoided for calculating likely heat as it is only emitted in short bursts.

The sensible heat output therefore should be based on the total CPO output divided by motor and battery efficiency (estimated at 90-95% efficient for motor, battery).

However, you still need to apply an average power utilisation factor to this – typically I use only 30% for ramp haul diesel trucks (uphill, downhill, rest average), which conservatively could be used for electric trucks too.

In fact, because trucks have regenerative braking, and a more efficient power train, it is likely to be a lower power utilisation factor - maybe in the order of 20%-25% or even less.

As a worked example, if the total CPO of a truck is 500kW, the calculated average Sensible Heat to include in Ventsim can be 500kW / 95% x 25% =  132kW.  Ensure this is entered as Sensible Heat, not Diesel Motor Power in Ventsim.  A future feature in Ventsim may have a specific BEV feature for this entry.

One last consideration, is McPherson states that moisture generated by (diesel) mobile machines has been measured at several times the theoretical moisture emitted from the exhaust, likely due to the moist loads carried by many machines, and the accelerated evaporation of moisture around the machine activity.  There is no reason to think BEV's would be any different.  Therefore, placing all heat as Sensible Heat only may produce overly dry and hot air temperature results.  Entering the heat as 50% Sensible and 50% Latent heat, or by adding 20ml/sec of moisture per 100kW of Sensible heat will inject some moisture into the model to produce more realistic results, although this amount should be ideally calibrated against measurements.  For the above example, the 132kW could be entered at 66kW Sensible + 66kW Latent, or alternatively 132kW Sensible + 26ml/sec of moisture (which the simulation will evaporate and convert from sensible to latent)

When simulated, you'll find significantly less heat and temperature increase should be generated by BEV for similar engine outputs.  Ultimately, BEV's have motor outputs 90-95% efficient (plus downhill and braking regeneration), versus only 30-35% efficiency for diesel engines.

 

[Martin Griffith]

Just a note on the above, we have now added the BEV heat source to the Heat Presets, as Craig foreshadowed at the end of the 5th paragraph of his post. This means that following the example above, for a total CPO of a truck of 500 kW with 25% utilisation, you can enter this as follows in the Heat Presets: 

 

Then there are 2 settings in Settings | Simulation | Heat which set the electric vehicle efficiency and also set the breakup of the vehicle heat output into sensible and latent heat (default setting is 50:50) as explained in the 6th paragraph of Craig's post.  

 

Any feedback, please let us know!

[Martin Griffith]

more information here: 

 

[Bruno Mota]

Thanks for the instructions on the BEV fleet.

[Adrian Halim]

Hi,

I notice that in Ventsim Design 6's heat calculator, the latent heat from the load carried by the BEV is included based on the latent heat factor in the settings. However, if the moisture is inputted as a separate heat source (i.e. wet broken rock carried by the BEV that is set as linear heat source or an activity track) then I should set the latent heat factor as zero, right? Otherwise, the moisture is simulated twice. Am I correct?

 

[Martin Griffith]

Hi Adrian, there is not a clear answer on this. The latent heat factor attempts to account for the evaporation of water around the vehicle, from the airway surfaces, and from any moisture sources on the vehicle. Much the same way as for a diesel vehicle, but without the moisture from the diesel combustion (see McPherson section 15.3.2.2)

If you know the moisture addition from the load and you add it as a separate heat source, then you could turn down the latent heat factor. But not to zero, as the latent heat factor is accounting for other moisture interactions around the vehicle as well. 

However, how much moisture exactly you should be adding, as well as where specifically it is coming from, is not known for certain and is an area of interest for us and I imagine many people looking at underground BEV use. 

On mine models, this sort of factor (like the rock wetness fraction) is something that can be tuned to match observed dry bulb temperatures.

 

[Adrian Halim]

Hi Martin,

I see, but how much should I reduce the latent heat factor? If the airway surfaces are dry, surely zero is a reasonable, right? Because I don't see other moisture sources on the BEV in this scenario. 

Also, how are these moisture interactions simulated with diesel machines? I don't see "Diesel machine latent heat factor" in the settings. There is "Water to Diesel ratio", but my understanding is that it only includes moisture produced by diesel fuel combustion.

I'm trying to simulate an apple-to-apple comparison between heat emitted by a BEV and its diesel version, so this is why I'm asking this question. It seems that the BEV's latent heat is overestimated from what I've read in this discussion, which is not the right way to do this simulation.

There are few field measurements of heat emitted by a BEV and its diesel version / counterpart that have been done in Sweden and Canada. Unfortunately none of them measured the moisture interactions. The one done in Sweden was carried out by me and my colleagues in Epiroc where we compared Epiroc's BEVs and their diesel version at Epiroc's Kvarntorp test mine in Sweden. Unfortunately we could not measure heat emitted by both versions since the mine strata is cold and absorbs most of the heat emitted by the machines (no temperature change observed when the BEVs were running and only little temperature increase was observed when the diesels were running). Therefore, we estimated the heat output by measuring the power consumed by both versions. Attached is a paper that shows results of this trial, which was presented at a conference in Stockholm last March. It shows that the ratio between power consumption (heat output) of the BEVs and that of their diesel version is lower than the published theoretical values. The trials done in Canada also found similar results.

10 BT_Adrianus_batterimaskiner_SIMS_NEXGENSIMS v2.pdf

[Martin Griffith]

The software has different settings for handling the sensible to latent ratio from diesel vehicles and BEV's and that is something we ideally should make consistent. We hear you on that, the difficulty would be to come up with a BEV equivalent of litres of water per litre of diesel. We will think more on this, it could involve changing the original diesel setting.

As is, the setting "Simulation | Heat | Water To Diesel Ratio" for diesel accounts for both the combustion products and the other associated moisture production from the machine and surrounding wet environment, as described in McPherson section 15.3.2.2. If you are operating your diesel vehicle in a very dry environment, it would stand to reason to reduce this ratio. 

In your case of the BEV where you are separately modelling the moisture from the load and you seem to be operating in a very dry environment, you could reduce the setting "Simulation | Heat | Electric Vehicle Latent Heat Factor" to a lower value; reducing to zero would imply there is no moisture in the surrounding environment, nor involved in any of the other mechanical processes on the vehicle.

To do a detailed investigation and comparison in Ventsim, you could work out the water-to-diesel ratio for diesel minus the moisture from the fuel (by subtracting 1.1 litres/litre from the ratio), work out the equivalent kW of latent heat per kW of diesel power attributable to the non-fuel combustion moisture evaporation, then apply the BEV latent heat factor percentage to match this for your BEV. This should work, however, with the assumption that the non-combustion moisture production is the same between BEV's and diesel vehicles (which may not be correct because of regenerative braking).

Thanks for the link to the paper, I think the results of such work are getting out there and users are setting lower utilisation factors on BEV's than they would use on equivalent diesel machines to reflect the lower power consumption from regenerative braking.

[Martin Griffith]

As discussed at the recent Ventsim User Conference, this is a developing field and the way that Ventsim handles this is not set in stone. 

Thinking more on this, one possible resolution is:

==============================

Current

Water to Diesel Ratio = 6.0 litres/litres

Electric Vehicle Latent Heat Factor = 50%

================================

Proposed

Water to Diesel Ratio = 6.0 litres/litres

Electric Vehicle Latent Heat Factor = 50%

Water to Diesel Stoichiometric Combustion Ratio = 1.1 litres/litres

Vehicle Work Moisture Production = X (litres/s)/kW output

==============================

Both diesel and electric vehicles would use the Vehicle Work Moisture Production setting, while diesel would add 1.1 litres more water per litre of diesel used, to account for combustion. This would allow an easier comparison, and mean that you wouldn't have to do the calculation yourself that I described in the second last paragraph of my previous post.

When opening old models we could generate the new setting based on the existing water to diesel ratio setting. The old Water to Diesel ratio setting could be left there as well and changing it would recalculate the Vehicle Work Moisture Production setting.

A problem is the assumption that the non-combustion related moisture production mechanisms on the vehicles (ie the Vehicle Work Moisture Production) are the same for electric and diesel, per unit of work output done. Which they might not be if you consider that the BEV is using regenerative braking and we may have a lower utilisation factor on the BEV in our model to reflect this, which could be a factor. Not sure of the answer there, any advice or suggestions welcome!

[Adrian Halim]

Thanks Martin. I wanted to attend the user conference but it clashed with my teaching schedule. Your proposal needs to be investigated further since I strongly believe that the "vehicle work moisture production" would not be the same for BEVs and diesels. The other reason beside the regenerative braking of BEVs is that the temperatures of electric motors and battery packs are much lower than the temperature of diesel engines. Large BEVs use liquid cooling system for their motor and battery pack, but since their temperature is much lower than diesel engines, the amount of evaporated coolant is significantly less than that in diesel cooling system. Probably some small BEVs use air-cooled motor and liquid-cooled battery pack, means that the amount of evaporated coolant is even lower in these vehicles. So perhaps the "vehicle work moisture production" should be split into one for diesels and one for BEVs. Field measurements must be done in order to quantify the amount of evaporated coolant in UG mine BEVs.

[Martin Griffith]

Thanks Adrian, yes, agree with your point about the lower temperature overall of the vehicle affecting the moisture production per unit of work output. Perhaps it would be more equivalent between the two types if formulated per unit of heat output, rather than work output.

We are not 100% sure at the moment so for now makes sense to retain separate settings between diesel and electric vehicles, and to keep an eye on research as it develops.

[Adrian Halim]

There is still a lot of uncertainties in input data for BEV heat simulations. For my case, I reckon that I'll use 1.1 as Water to Diesel Ratio and zero for Electric Vehicle Latent Heat Factor. Until field measurement results are available, I wouldn't dare to guesstimate vehicle work moisture production for both diesels and BEVs operating in dry conditions. In my case, I reckon that most of the extra moisture come from the broken rock, which is already inputted separately. 

Also, on the list of preset "Heat Moisture Diesel", using Craig's example above, if I put (500 kW x 0,25 = 125 kW) as the Electric Vehicle Power and 1 as the Utilisation factor, will it produce the same heat input as what is shown above (500 kW as the Electric Vehicle Power and 0.25 as the Utilisation factor)? I saw other users do this way.

Beside uncertainties in input data for BEV heat simulations, the same is happening for BEV fire simulations. I responded to Florian's post in UVC discussion group. Perhaps you've read it. It's good that Ventsim Design 6 has features to do heat and fire simulations for BEVs, but the users must be cautioned about the uncertainties in the preset input data for these simulations. There is still a lot of work to do to make these simulations produce accurate results. I'm hoping that I'll get opportunities to do field trials and tests to obtain more accurate input data than the preset values.

[Martin Griffith]

"Also, on the list of preset "Heat Moisture Diesel", using Craig's example above, if I put (500 kW x 0,25 = 125 kW) as the Electric Vehicle Power and 1 as the Utilisation factor, will it produce the same heat input as what is shown above (500 kW as the Electric Vehicle Power and 0.25 as the Utilisation factor)? I saw other users do this way."

Yes, that is how it works. 

[Adrian Halim]

Thanks Martin. I need you to check my model since the WB difference between BEVs and diesels is less than what I expected. I've sent an email to you explaining what I've done. Hope that it got through since the model is 21 MB.