Dynamic performance of reformate-fed PEMFC hybrid bus

4
...Mr.n1M lllIIIMlIIW o l JU:. 1999 FEATURE Dynamic performance of reformate-fed PEMFC hybrid bus By Louise Potter and Jessica Reinkingh, Johnson Matthey Pic, UK/USA Drlve-cyde characteristics Drlve-eyde I Drive-cycle II Total energy demand, kWh 7.02 63.87 Net energy consumption," kWh 6.57 60.23 Maximum power demand, kW 107.23 109.0 Average power demand, kW 26.2 20.3 Drive-cycle time (hr:min:sec) 0:15:03 2:58:28 Maximum speed, krnIh 40.75 52.23 Average speed, kmlh 16.60 15.33 *Dlfference due to regenerative braking. This article presents results of a study to assess system design options for proton• exchange membrane (PEM) fuel cell buses. During the study, the dynamic performance of a reformate-fed fuel cell system in a hybrid bus was simulated, based on data measured during operation of an existing bus using a hybrid internal combustion engine (ICE). Buses will be the earliest widespread transportation application of fuel cells. The introduction of fuel cell buses will reduce the levels of urban pollution, and be a very visible example of the advantages of fuel cell vehicles. Already, hydrogen fuelled fuel cell buses are undergoing trials in Chicago and Vancouver. However. there are uncertainties over the hydrogen infrastructure and cost. This has spurred the interest in on-board hydrogen. Reforming fuels such as methanol. ethanol. natural gas and gasoline can produce hydrogen • of these. methanol is considered the "easiest" liquid fuel to reform. Researchers at Georgetown University! 11 have developed and demonstrated the first full-size bus operating on a phosphoric acid fuel cell (PAFC), including an on-board hydrogen generator with methanol as the fuel. This methanol-fueled bus is a hybrid vehicle comprising a 100 kW PAFC system and 180 Ah NiCd battery stack. Georgetown is also developing a hybrid bus based on a PEMFC system. This bus is expected to be operational by the end of 1999. As a developer of catalysed fuel cell com• ponents for the stack and fuel processor. Johnson Matthey conducted a stud y l 2 1 to assess system designs for PEM fuel cell buses. This study addresses issues including the needs of the bus industry. the choice of fuel. fuel availability. fuel processing technology. the power requirements for the fuel cell system, and the dynamic res• ponse of a methanol reformate-fed hybrid system. This article concentrates on the latter issue. Drive-cycle power measurements The fuel cell system model was based on data from an operating petrol hybrid bus. built by Thoreb and Scania (see Table 1). Battery 10 NiCd batteries, connected in series Charge/discharge potential: 405-324 V Energy when fully charged: 25.9 kWh Rated battery capacity: 80 Ah Table 1. Hybtld bus powertram specification Table 2 Average values over bus dtlve-cycle The power flows in this bus were evaluated over two separate urban drive-cycles in Stockholm, distinguishing between total power demand. and the power supplied by the ICE and batteries. The average characteristics of the drive-cycles are shown in Table 2. Detailed data for one drive-cycle are given in Figure 1. In comparison to this Stockholm route. the most representative standardised test cycle for inner city buses, the Braunschweig cycle, is slightly more demanding (average power -30 kW. average speed -23 km/h). ICE 2.3 litre Saab engine Operates in 3 modes to optimise efficiency: 3500 rev/min if speed>7 km/h 2750 rev/min if speed <7 kmlh •0 rev/min, powered by batteries, when driving through "emission-free" zone ---------------------------'5' Fuel Cells Bulletin No.1 0 \.:J

Transcript of Dynamic performance of reformate-fed PEMFC hybrid bus

...Mr.n1M lllIIIMlIIWbY1ia:~

o l JU:. 1999

FEATURE

Dynamic performance ofreformate-fed PEMFChybrid busBy Louise Potter and Jessica Reinkingh, Johnson Matthey Pic, UK/USA

Drlve-cyde characteristics Drlve-eyde I Drive-cycle II

Total energy demand, kWh 7.02 63.87

Net energy consumption," kWh 6.57 60.23

Maximum power demand, kW 107.23 109.0

Average power demand, kW 26.2 20.3

Drive-cycle time (hr:min:sec) 0:15:03 2:58:28

Maximum speed, krnIh 40.75 52.23

Average speed, kmlh 16.60 15.33

*Dlfference due to regenerative braking.

This article presents results of a study to assess system design options for proton•exchange membrane (PEM) fuel cell buses. During the study, the dynamicperformance of a reformate-fed fuel cell system in a hybrid bus was simulated,based on data measured during operation of an existing bus using a hybridinternal combustion engine (ICE).

Buses will be the earliest widespread

transportation application of fuel cells. The

introduction of fuel cell buses will reduce the

levels of urban pollution, and be a very visible

example of the advantages of fuel cell vehicles.

Already, hydrogen fuelled fuel cell buses are

undergoing trials in Chicago and Vancouver.

However. there are uncertainties over the

hydrogen infrastructure and cost. This has

spurred the interest in on-board hydrogen.

Reforming fuels such as methanol. ethanol.

natural gas and gasoline can produce hydrogen •

of these. methanol is considered the "easiest"

liquid fuel to reform.Researchers at Georgetown University! 11 have

developed and demonstrated the first full-size

bus operating on a phosphoric acid fuel cell

(PAFC), including an on-board hydrogen

generator with methanol as the fuel. This

methanol-fueled bus is a hybrid vehicle

comprising a 100 kW PAFC system and 180 Ah

NiCd battery stack. Georgetown is also

developing a hybrid bus based on a PEMFC

system. This bus is expected to be operational by

the end of 1999.As a developer of catalysed fuel cell com•

ponents for the stack and fuel processor. Johnson

Matthey conducted a studyl21 to assess system

designs for PEM fuel cell buses. This study

addresses issues including the needs of the bus

industry. the choice of fuel. fuel availability. fuel

processing technology. the power requirements

for the fuel cell system, and the dynamic res•

ponse ofa methanol reformate-fed hybrid system.

This article concentrates on the latter issue.

Drive-cycle powermeasurementsThe fuel cell system model was based on data

from an operating petrol hybrid bus. built by

Thoreb and Scania (see Table 1).

Battery

10 NiCd batteries, connected in series

Charge/discharge potential: 405-324 V

Energy when fully charged: 25.9 kWh

Rated battery capacity: 80 Ah

Table 1. Hybtld bus powertram specification

Table 2 Average values over bus dtlve-cycle

The power flows in this bus were evaluated

over two separate urban drive-cycles in

Stockholm, distinguishing between total

power demand. and the power supplied by theICE and batteries. The average characteristics

of the drive-cycles are shown in Table 2.

Detailed data for one drive-cycle are given in

Figure 1.In comparison to this Stockholm route. the

most representative standardised test cycle for

inner city buses, the Braunschweig cycle, is

slightly more demanding (average power

-30 kW. average speed -23 km/h).

ICE

2.3 litre Saab engine

Operates in 3 modes to optimise efficiency:

• 3500 rev/min if speed>7km/h

• 2750 rev/min if speed <7 kmlh

• 0 rev/min, powered by batteries, whendriving through "emission-free" zone

---------------------------'5'Fuel Cells Bulletin No.1 0 \.:J

FEATURE

Table 3 Input variables considered In fuel cell system (FCS) power supply scenarios.

change in power demand of the bus. With

a minimum value of 5 kW, the maximum

power output of 50 kW is reached for less than

I% of the drive-cycle. Furthermore. under these

operating conditions the system would never

need to switch off, and actually loses battery

power. Figure 2 shows the power supply by the

fuel cell system and the battery, as well as the

total power demand and the battery state-of•charge (SOC), for a slow-response fuel cellsystem.

With an acceleration of 1 kW Is the fuel

cell system is unable to follow the load.

So, operating the FC system at a constant

load would be more sensible. The maximum

power of [he Fe system should be slightly

higher than the average power demand, to

ensure the batteries will not be drained. Figure 3shows the results with a constant fuel cell systemload.

In practice the average power demand Over

a drive-cycle would not be exactly known, and

in fact, even for a particular route, will

change slightly on a daily basis depending on

the outside conditions and number of

passengers. Hence, it would be more Hexible to

build a powertrain with a fuel cell system

operating with a semi-constant load. In this

mode, the fuel cell system power output would

vary between a maximum and minimum,

dictated by the batteries rather than the total

power demand. However, such a strategy can

only be implemented if the relevant battery

parameters can be measured and translated

reliably into a value for the battery state-of•charge.

95@~

90 ~

80

85

100

0:14

0:14

Battery SOC

I

0:12

0:12

0:10

0:10

0:08

0:08

ICE Power

Minimum state-of-charge SOCmin (~FCS on), %

Minimum power step-size of FCS (.s), kW/s

Maximum state-of-charge SOC""" hFCS off), %

Total battery energy at 100% SOC (capacity), kWh

0:06

0:060:04

Battery Power

Total Power Demand

~

0:02

- Total Power Demand - Battery Power -ICE Power - Battery SOC

105

100

-1000:00

-50

"C 50 .. _._ ..~.- --

~ ....(/)~~=.a:I 0

0:00 0:02 0:04

Acceleration of FCS (A), kW/s

Minimum net FCS power (Min), kW

Maximum net FCS power (Max), kW

Deceleration of FCS (0), kW/s

Figure 1 Power demand/supply In present hybrtd bus (drive cycle I).

~-"'-"'/'/-'>f- I-'~ f.' I"'" 100

~ 50 1\l,..- 1/\.,

~v· 17 ii'

"C /

~95 ......

c:1/ ~

co~

1/E h.Q) ~ ~",

0 0

V...

l~,j \1 ~ Ii,f ~...

J I\iI 90 I~ I

Q.

i·50

r85I 11 ~

-100 800:00 0:02 0:04 0:06 0:08 0:10 0:12 0:14

Drive Cycle Time [h:min]

-FCS PowerFuel cell system dynamicrequirementsThe fuel cell system was modelled as a function

of input variables, allowing the gap between the

fuel cell system and the ICE to be assessed

(Table 3). This gap analysis forms the basis of the

fuel cell hybrid bus power requirements. eval•

uating power demands on the batteries for the

various fuel cell system power supply scenarios.

In addition, a number of output parameters

for the ICE generator and the fuel cell system

were calculated. The most important ones are

summarised in Table 4.In the simulations, the extremes of the system

were explored (very slow and very fast response),

followed by more realistic intermediate values.

Slow fuel cell system response

- Power Demanded

100

- Battery Power - Battery SOC.. -~ ._.__._----------_.-_._~ 105

It was found that with an acceleration of 1 kW/s,

the fuel cell system is very slow compared to the

Figure 2. Simulated power supply In fuel cell system bus (A = 1 kW/s, D = -1 kW/s. Max = 50 kW,Min = 5 kW).

®-----------------------~~--Fuel Cells Bulletin No. iO

FEATURE

The maximum fuel cell system power supply

needs to be increased to cover the full load, in

this case 110 kW. The minimum power

requirement should be zero. As the fuel cell

system response rate is increased, the reliance on

the batteries is reduced. figure 4 shows the

results for the fuel cell system with an accel•

eration capability of 55 kW/s. It can be seen that

there are two very small spikes where the battery

delivered energy (0 ..1% of the total energy

demand). If the fuel cell system response rate

decreased from 55 kWIs, more battery spikes

appeared. So, to remove the batteries completely,

the fuel cell system response rate needs to be at

least 55 kW/s for the given drive-cycle.

Fast response: Fuel cell system-onlybus

80

100

~

95Lgen~

90 ~~

85

0:140:120:04 0:06 0:08 0:10

Drive Cycle Time [h:min]0:02

----

-~f../ k/' '",~

/1\ I"

~\::1\ ~//!

l Po~

t-'r---/ ~ .r I~

rl ili~It' I

J II

1~i

1It

o

50

-50

-1000:00

- Power Demanded - Battery Power - FCS Power - Battery SOC

105

100

Figure 3 Fuel cell system constant power supply scenario (A = 1 kW/s. D =-A. Max =27 kW,Min =27 kW) Intermediate response

Having investigated a very slow and a fast fuel

cell system response, the power demand and

supply was also studied for 2, 5, 10 and

20 kW Is. The faster the fuel cell system

responds, the more load-following it becomes.

The batteries are used less, and consequently the

required battery capaciry decreases. However, the

maximum fuel cell system power output must

increase accordingly. It was found that

satisfactory performance could be achieved with

an acceleration rate of 5 kWis. Figure 5 shows

the results, and includes a .1-minute fuel cell

system start-up time, where the system is heated

by burning methanol. It was found that a start•

up time of 3 min should cause no problems for

the batteries. Idling the bus for longer before

driving off will significantly reduce the impact

on the batteries.

Hybrid system sizeThe objective of replacing the ICE in the current

hybrid bus was to create an efficienr zero•

emission vehicle. However, the vehicle should

still have a respectable driving range, without

giving up space or increasing the weight. In

Table 5 the system weight and volume are

estimated as a function of the fuel cell system's

acceleration. The hybrid fuel cell system

powertrain is estimated to be slightly larger than

the current hybrid ICE powertrain. However, as

the fuel cell system response rate increases, less

battery is required, and the weight of the system

drops.

Conclusions

100

~

95Lgen~

90 Q)

tiCD

85

80

0:14

- Battery SOC

105

Fraction of drive-cycle at zero-power, %

Fraction of drive-cycle at Max power, %

Fraction of drive-cycle at Min power, %

Maximum power deceleration, kW/s

Average power deceleration. kW/s

Average power supply, kW

Average power acceleration, kW/s

Maximum power acceleration, kW/s

Total energy supply, kWh

ICE or FCS

- Battery Power - FCS Power

0:04 0:06 0:08 0:10 0:12

Drive Cycle Time [h:min]

-

~'"

~ " ~

,.......... ~ ... it J-1 ij

50

o

-50

100

·1000:00 0:02

- Power Demanded

Figure 4 Fuel cell system-only, completely load-followlIlg (A = 55 kWls, D =-A, Max = 110 kW, Min =o kW, 5 = 0 5 kW/s)

Net energy supply, kWh

Average power charge rate, kW

.' Minimum SOC, %

"'~ power charge rate, kW

-MaxImum power discharge rate, %/s

:.Maxlmum power discharge rate, kWJ

Average power discharge rate, %/s

. ·Average power discharge rate, kW

Tabl<> 4 Calculated output values for battenes and ICE or fuel cell system The simulations have shown that in the

hybrid bus studied, the battery capacity is

---.......-------------------------([)Fuel Cells Bulletin No.1 0

FEATURE

FCS/ICE acceleration Diesel ICE Hybrid ICE 1 2 5 10 20 55Number of batteries 10 10 10 8 6 5 0

Maximum FCSIICE power 50 35 37 45 55 60 110

Battery volume 1000 1000 1000 800 600 500 0

FCS/ICE volume 2200 200 700 740 900 1100 1200 2200

Total Volume 2200 1200 1700 1740 1700 1700 1700 2200,

Battery weight 1000 1000 1000 800 600 500 0

FCSIICE weight 1200 400 350 370 450 550 600 1100

Total weight 1200 1400 1350 1370 1250 1150 1100 1100

The fuel cell system's specific power is 0.1 kW/kg (0.05 kW/I).

Table 5 Estimated hybrid system weight and volume,

- Power Demanded ~ Battery Power - FCS Power - Battery SOC------------ 105

Figure 5 Power flow in fuel cell system hybrid bus during drive cycle I with 3 min FC system start-up,using heat generated by methanol combustion in anode off-gas burner (A =5 kW/s, 0 =-5 kW/s, Max=50 kW, Min =15 kW, 8 batteries).

hybrid. For a more responsive fuel cell system,the overall weight of the powertrain is expectedto decrease, as less battery capacity is required.

ReferencesI. Presentation by Jim Larkins, ProgrammeManager, Georgetown University Fuel Cell Tran•sit Bus Program, at IQPC Fuel Cell TechnologyConference, London, UK, September 1998.

2. L.c. Potter, J.G. Reinkingh: "SPFC busdesign studies". ETSU report F/02/00134/REp,May 1999.

AcknowledgmentsJohnson Matthey acknowledges the sUppOrt ofthe Energy Technology Support Unit (ETSU) ofthe UK Department of Trade & Industry, whosupported the fuel cell bus studies undercontract F102/00134, Johnson Matthey is alsograteful to the input, data and support fromKristian Julen and Mattias Forsen at Thoteb AB,Vasrra Frolunda, Sweden,85

800:140:120:100:080:060:040:02

FCS start-up time.. 3 minutes·

L- ~__.. --'

Drive Cycle Time [h:min]

100

-100

0:00

sufficient for a direct replacement of the ICEwith a methanol reformer fuel cell system with adynamic response of 1-2 kW/s. However,at 5-10 kW/s, the battery capacity couldbe reduced by 20-40%. The maximumpower for such a system should be about

45-55 kWe. It was found that for a dynamicresponse of5 kW/s, a 50 kW (net) fuel cell systemwould be reasonable for the given drive-cycle.

It is expected that such a powertrain would beof similar volume and weight to a conventionaldiesel ICE, and lighter than a diesel/battery

For further information. contact: louise Potter.Johnson Matthey Technology Centre, Blount's Court,

Sonning Common, Reading RG4 9NH, UK. Tel: + 44 1189242123, Fax: + 44 118 924 2360, Email: pottelc@

matthey.com

Editorial ContributionsThe editor is pleased to consider news items and articles for publication in the newsletter which discuss technical aspects ofall typeS of fuel cellsand related technologies. Submissions concerning developments in and applications of fuel cell technology are particularly welcomed. Ideally.articles should be approximately 2000 words long with illustrations/photographs as appropriate. News items should be between 100 and 500words.

Ifyou art interested then please contact: Steve Barmt, Fuel Cells Bulletin, ElsevierAdvanced Technology, PO Box 150, KitJ/ington, Oxford. 0X51AS.UK.Tel: +441865 843239;ftJX: +441865853160; e-mail: s. [email protected]

0------------------------~~--Fuel Cells Bulletin No.iO