Hydrogen Fuel Cell Buses: Modelling and Analysing Suitability froman Operator and Environmental Perspective

Doyle, D., Harris, A., Douglas, L., Early, J., & Best, R. (2020). Hydrogen Fuel Cell Buses: Modelling andAnalysing Suitability from an Operator and Environmental Perspective. In 2020-04-14 Hydrogen Fuel CellBuses: Modelling and Analysing Suitability from an Operational and Environmental Perspective 2020-01-1172(Technical Papers ). SAE International. https://www.sae.org/publications/technical-papers/content/2020-01-1172/Published in:2020-04-14 Hydrogen Fuel Cell Buses: Modelling and Analysing Suitability from an Operational andEnvironmental Perspective 2020-01-1172

Document Version:Peer reviewed version

Queen's University Belfast - Research Portal:Link to publication record in Queen's University Belfast Research Portal

Publisher rightsCopyright 2020 SAE International. This work is made available online in accordance with the publisher’s policies. Please refer to anyapplicable terms of use of the publisher.

General rightsCopyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or othercopyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associatedwith these rights.

Take down policyThe Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made toensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in theResearch Portal that you believe breaches copyright or violates any law, please contact openaccess@qub.ac.uk.

Download date:25. Dec. 2021

Page 1 of 11

01/11/2016

2020-01-1172

Hydrogen Fuel Cell Buses: Modelling and Analysing Suitability from an

Operational and Environmental Perspective

Author, co-author (Do NOT enter this information. It will be pulled from participant tab in

MyTechZone) Affiliation (Do NOT enter this information. It will be pulled from participant tab in MyTechZone)

Abstract

In response to the need to decrease greenhouse gas emissions, the current trend in the transport sector is a greater focus on alternative powertrains. More recent air quality concerns have seen controlled zero emission zones within urban areas. As a result, there is growing interest in hydrogen fuel cell electric buses (FCEBs) as a zero local emission vehicle with superior range, operational flexibility and refuelling time than other clean alternatives e.g. battery electric buses

(BEBs). This paper details the performance and suitability analysis of a proposed Wrightbus FCEB, using a quasistatic backwards-facing Simulink powertrain model. The model is validated against existing prototype vehicle data (Mk1), allowing it to be further leveraged for predictions of an advanced future production vehicle (Mk2) with next generation motors and fuel cell stack. The modelled outputs are used for a comparison of the FCEB performance to an equivalent BEB on industry standard drive cycles, as well as several real bus routes

generated through data logging activities. The suitability of FCEB vs BEB from an operator usage perspective is thus analysed in different use cases, with case studies from the UK and Chile. Both single-deck and double-deck vehicle types are considered. Modelled FCEB and BEB outputs are further utilised in a comparative well-to-wheel assessment, highlighting that the relative suitability of each from an environmental perspective is sensitive to geographical and fuel production method pathways. The paper concludes with a discussion

of the environmental and societal benefits of deploying hydrogen buses to reduce local health damaging pollutants and alleviate energy security concerns via the introduction of a feasible and sustainable transport alternative.

Introduction

Global commitments to decrease greenhouse gas emissions have led to a shift to alternative powertrains in the transport sector. In addition to this, stricter controls on air quality within cities has seen the introduction of zero emission zones, requiring vehicles with full zero emission capabilities. As a result, there is growing interest in hydrogen fuel cell electric buses (FCEBs) as a zero local emission vehicle with superior range, operational flexibility and refuelling time than other clean alternatives e.g. battery electric buses (BEBs). This is illustrated in increased investment through projects such as

JIVE/JIVE2, which are deploying nearly 300 FCEBs and refuelling infrastructure in Europe by the early 2020s [1]. Currently, FCEBs and BEBs make up 2 % and 19 % of the European bus market respectively, with those figures predicted to rise to 10 % and 52 % by

2030 [2]. In the UK, Aberdeen City Council are working towards a hydrogen economy through the H2 Aberdeen initiative, who join London in placing orders for the world’s first double-deck FCEBs to

add to their existing single-deck fleet [3–5]. Chile and Norway are pursuing the production of hydrogen from renewable sources, with some predictions for future prices reaching as low as $1.60/kg for Chile and $2.20/kg for Norway [6]. Australia and Saudi Arabia have also been identified as potential producers [6].

The objective of this paper is to initially establish the modelled capabilities of a future production Wrightbus FCEB with recent fuel cell and battery technology. Next, the suitability of the FCEB and

corresponding BEB as local zero emission options are compared, along with a comparison of single and double-deck variants. Case studies from the UK and Chile are considered. The paper will compare the vehicles from two perspectives:

1. The bus operator – considering vehicle range, passenger carrying capabilities, operational cost/efficiency and level of service.

2. Environmental – considering a well-to-wheel assessment

based on geographical location and fuel production method.

Mahmoud et al [7] comprehensively review literature on alternative powertrain buses considering operational, environmental and economic factors. From an economic perspective, the total cost of ownership of FCEBs and BEBs was widely uncertain in literature. Figures used throughout the review are taken from other studies (simulation or operational data) and often averaged for fairness. Whilst overall fair, it may not represent a direct comparison in all

circumstances, depending on the type of study used. In addition, advances in relevant technology since the review may set new benchmarks influencing the operational and environmental aspects of the study.

More recently, Correa et al [8] compare alternative powertrain buses from an energy and environmental perspective using characteristic vehicle parameters from the simulation software, ADVISOR. For the drive cycles considered, the road is assumed to be flat. Several recent studies consider an environmental assessment of FCEBs, for

operation in Taiwan and the United States [9,10]. Neither study also analyses BEBs as a comparative zero emission vehicle and also do not consider operational requirements.

This paper adds to the current literature firstly by demonstrating the capabilities of a future production FCEB availing of the most recent

Page 2 of 11

01/11/2016

fuel cell and battery technologies available. Technology in this area is developing rapidly, and this work will model predicted performance with state-of-the-art production components. Working with Wrightbus to develop a model which is representative of a production intent vehicle powertrain avoids using generic parameters. This work

is of a similar process to existing work at Queen’s University Belfast in conjunction with Wrightbus [11–15]. All powertrain parameters and control logic are known and implemented. Additionally, the hydrogen consumption of the model is validated with a prototype vehicle.

Whilst many previous studies investigate the environmental impacts of FCEBs and compare with BEBs, fewer also incorporate an operational perspective. This work considers both, and is the first

opportunity for a comparison of the two vehicle types designed with production intent on a common, modular platform with shared components. Adding to this, the difference between single and double-decks is considered. Real bus routes are simulated for which vehicle utilisation is known, to allow consideration of the level of service possible. Finally, the use of real bus routes allows the influence of the road gradient on the overall consumption to be accounted for.

Methodology

This section gives an overview of the vehicles modelled, modelling method, validation efforts and the drive cycles considered.

Two different local zero emission bus technologies are considered: FCEBs and BEBs. For each, single-deck and double-deck options are

modelled. The first model is of a prototype Wrightbus FCEB (Mk1) [17], for which detailed vehicle parameters and driving test data exists for validation. This validated model is then leveraged to predict the performance of an intended future production FCEB (Mk2), which is used for the comparative analysis to the BEB. Mk2 avails of next generation technology compared to Mk1, available for production use now. As well as a different tractive motors, a new fuel cell offering improvements in size, reliability and total cost of ownership is used [18]. Battery chemistry changes from lithium-

titanate (LTO) to lithium-nickel-manganese-cobalt-oxide (NMC) which is less expensive and offers a higher specific energy. Changes to the power electronics mean that despite the decrease in specific power of the NMC cells, the charging current limit of the battery increases, allowing greater energy recuperation through regenerative braking.

The BEB is an overnight charging variant from the same manufacturer and based on the same 10.6 m platform as the FCEB

Mk2. Thus, the motor, axle, chassis and body are consistent between vehicles. The battery also uses the same NMC cells as FCEB Mk2, with a larger quantity to provide the additional capacity required in the absence of a range extender. The increase in battery mass from FCEB to BEB is offset by the mass of the fuel cell, support structure and auxiliary systems, as well as the hydrogen tanks, associated fittings and enclosures/crash-worthy structure. The commonality of components in each vehicle provide an opportunity to compare

FCEBs and BEBs consistently and fairly. Table 1 lists the vehicle parameters used throughout. As the Mk1 FCEB is used only for model validation and not for the comparative analysis, it has been omitted.

Table 1. Vehicle parameters.

FCEB (Mk2) BEB

Singe-

deck

Double-

deck

Single-

deck

Double-

deck

GVW (kg) 18,000 19,500 18,000 19,500

Passenger Capacity 72 100 72 100

Battery Capacity

(kWh) 53.96 339.18

Hydrogen Capacity

@350 bar and 15ºC

(kg)

30.88 26.78 N/A N/A

Modelling

The vehicles are modelled as quasi-static backwards-facing powertrain models within Simulink/MATLAB. Drive cycles consisting of velocity and elevation profiles form the model input,

with power-flow working backwards from the required velocity/elevation change at each timestep to the battery. Lateral motion was neglected. This modelling methodology neglects transients and thus cannot be used for detailed vehicle control and response, but gives a good prediction of macro quantities such as energy and fuel usage. Component specifications and efficiency maps are implemented based on OEM datasheets and laboratory test data where available. Figures 1 and 2 show the vehicle architecture for the

FCEB and BEB. Both are identical with the exception of the fuel cell stack, and an associated brake resistor to dissipate excess current during significant regenerative braking events.

Figure 1. FCEB model schematic.

Fuel Cell

DC/DC

Auxiliaries

Inverter Battery Motor

Axl

e/

Dif

f

DC/DC

Brake Resistor

Page 3 of 11

01/11/2016

Figure 2. BEB model schematic.

Validation

Extensive validation of the models is difficult for several reasons. The BEB and FCEB Mk2 are future production vehicles and therefore there is no physical vehicle for testing. FCEB Mk1 is a

physical prototype of a double-decker, however only limited test data is available. Furthermore, the added uncertainty of the HVAC power usage during testing must be eliminated for validation of the model, and thus only test data with HVAC turned off can be used. Quantifying the HVAC load adds uncertainty as it is dependent on the weather during the test (ambient temperature, level of sunlight and humidity), initial cabin temperature, bus velocity profile, passenger loading/unloading and the frequency and amount of time

the bus doors were open; amongst other factors. This leaves one test run for the validation of predicted macro quantities with the known condition of zero HVAC load, listed in Table 2. This test can easily be repeated for future tests as the route (and all stops) is known, and the velocity profile of the test route was recorded. Repeatability of the test can be achieved through either following the same route or recreation on a rolling road.

Table 2. FCEB Mk1 model predicted hydrogen consumption compared to

measured consumption during a test.

FCEB Mk1

Test

FCEB Mk1

Model

Prediction

Error

Hydrogen Usage

(kg/100km) 6.58 6.68 +1.52 %

Although the result is promising, it is only a validation of one route without HVAC, and only for the FCEB Mk1 vehicle. However, it is assumed that because the FCEB Mk2 model is derived from Mk1, that it will be similarly accurate. Additionally, the commonality of

components and the use of the same modelling techniques throughout the models minimises the opportunity for variation. Creation of the FCEB Mk1 model was for validation purposes only. For the remainder of this paper, any values attributed to the FCEB refer to the Mk2 vehicle, as it is on the same platform at the BEB.

Although there is no physical BEB for validation, the Office for Low Emission Vehicles/Low Carbon Vehicle Partnership in the UK operates a Low Emission Bus (LEB) scheme offering financial

incentives against the purchase of LEB accredited vehicles. Vehicles undergo a specified drive cycle (LUB cycle) and their emissions performance compared to a baseline set by diesel buses. Test certificates are published online along with the test conditions and

results, allowing them to be simulated within the model [19]. Table 3 compares the actual LEB test results for BEBs from other manufacturers with the values predicted by the model. Note that the simulated values represent predictions for the modelled Wrightbus BEB during an LUB drive cycle at the same vehicle mass as the LEB

test vehicle. Thus, they are not intended to validate the Wrigthbus BEB model specifically, but indicate that it is giving results in line with other BEBs. The two results show different levels of accuracy, which is to be expected as the powertrain parameters were not known and were thus simulated with the Wrighbus BEB powertrain. It is likely that when the modelled BEB performs the same test the accuracy will increase as the assumptions are reduced.

Table 3. BEB model predicted energy usage compared to measured usage

during LEB accreditation for other BEBs.

Vehicle Test Mass

(kg)

LEB Test

Energy

Usage

(kWh/km)

Model

Energy

Usage

(kWh/km)

Error

Volvo 7900

EV 13,471 0.847 0.789 -6.85 %

BYD eBUs

12m 13,868 0.831 0.802 -3.50 %

Drive Cycles

Several different drive cycles are used in this paper to illustrate the performance of the FCEB under various conditions. The first set of

drive cycles are industry standard and act as a performance benchmark to demonstrate the capabilities of the FCEB and BEB. The first is the LUB cycle used in the validation section of this report, which is the test for the LEB scheme. The LUB cycle is a modification of the previous standard cycle (MLTB) to include a rural phase in addition to the outer/inner city phases, typical of UK operation [20]. Next are the three synthetic “Standardised On-Road Test” (SORT) cycles developed by the UITP Bus Committee

specifically for comparing energy consumption between different buses [21]. SORT 1,2 and 3 represent heavy urban, urban and suburban respectively.

In addition to this, four drive cycles have been created from data logging of real bus routes to indicate real-world performance. The drive cycles consist of both the outbound and return journeys, therefore capturing the full characteristics of the route. Two relatively flat urban routes from London (UK), a moderately hilly suburban route from Santiago (Chile), and a steep ascent then descent in

Rancagua (Chile). This final route serves as transport for workers to and from a mine with an elevation change of almost 1500 m, and therefore represents a very challenging route for zero emissions capable vehicles. Table 4 lists the drive cycles with some metrics to characterise them. Relative positive acceleration (RPA) is a measure of the power required for all accelerations (ignoring decelerations) in the drive cycle, normalised by the cycle length [22]. It is essentially a measure of the aggressiveness of acceleration, and influences

fuel/energy consumption. Relative slope is the elevation range (maximum – minimum) normalised by the cycle length. Trips per day is a record of how many times the buses used for data logging were required to complete the route in one day to achieve the desired level of service.

Auxiliaries

Inverter Battery Motor

Axl

e/

Dif

f

DC/DC

Page 4 of 11

01/11/2016

Table 4. Summary of drive cycles created from data logging of real bus

routes.

RPA (m/s) Relative

Slope

Average

Speed

(km/h)

Stops per

km

Urban -

London 1 0.257 0.649e-3 19.2 2.76

Urban -

London 2 0.214 1.09e-3 14.7 4.46

Suburban -

Santiago 0.185 1.95e-3 25.6 2.56

Hilly -

Rancagua 0.111 15.9e-3 28.3 0.110

Results & Discussion

This section details the modelled vehicle performance on industry

standard drive cycles, followed by a comparison of each vehicle operating on real bus routes from both an operator and environmental perspective. Each of the industry standard and the real bus routes from Table 4 were simulated in the vehicle models. For each simulation, the vehicle was assumed to be at 50 % of it’s passenger capacity, and with no HVAC power requirement due to the difficulties in accurately quantifying it. All per-passenger values in this paper treat the vehicle as half-laden. Note that all figures given are predicted values from modelling efforts and may differ from

actual fuel/energy consumption.

Industry Standard Drive Cycles

Table 5 shows the simulated hydrogen/energy usage per km for each vehicle whilst driving on the four industry standard drive cycles. The results shown here are for demonstration of the vehicle’s capabilities under known operating conditions.

Table 5. FCEB and BEB hydrogen/energy usage for four industry standard

drive cycles.

FCEB Hydrogen Usage

(kg/100km)

BEB Energy Usage

(kWh/km)

Single-

deck

Double-

deck Singe-deck

Double-

deck

LUB 5.63 6.03 0.891 0.928

SORT 1 6.30 6.57 0.985 1.03

SORT 2 5.67 6.13 0.905 0.955

SORT 3 5.38 5.94 0.863 0.915

Average 5.74 6.17 0.911 0.956

To allow a direct comparison of the FCEB and BEB, the consumption values in Table 5 are converted to a maximum range for each vehicle and drive cycle (plotted in Figure 3), taking in to account the percentage of the battery capacity which is not used for battery life considerations.

Figure 3. Predicted vehicle range on four industry standard bus routes.

Real Bus Routes – Operational Perspective

Table 6 shows the simulated hydrogen/energy usage per km for each vehicle whilst driving on the four real bus routes. As a better representation of real-world conditions than the previous drive cycles, these are the results which will form the analysis for the remainder of the paper.

Table 6. FCEB and BEB hydrogen/energy usage for four real bus routes.

FCEB Hydrogen Usage

(kg/100km)

BEB Energy Usage

(kWh/km)

Single-

deck

Double-

deck Singe-deck

Double-

deck

London 1 6.95 7.49 1.08 1.12

London 2 6.11 6.30 0.948 0.988

Santiago 5.87 6.15 0.880 0.945

Rancagua 6.69 6.96 0.844 0.879

Average 6.41 6.73 0.938 0.983

Page 5 of 11

01/11/2016

Efficiency & Operating Cost

Naturally, the single-deck variants of each vehicle consume less hydrogen/energy per km than their equivalent double-decks due to their smaller mass and reduced drag – around 5 % on average. However, when expressed on a per-passenger basis, this trend is reversed. The FCEB single-deck is on average 32 % less efficient per passenger-km then the double-deck. Similarly, the BEB single-deck

is 33 % less efficient than the double-deck.

Using the average per-passenger values in Table 6, the sensitivity of the vehicle operating cost to the price of hydrogen/electricity can be determined. Figure 4 shows the cost delta incurred per 100-passenger-km by using a double-deck FCEB over a BEB. The cost delta is plotted as a function of the hydrogen and charging electricity prices (positive values indicate that the FCEB is more expensive to operate). Figure 5 shows the same analysis for the single-deck

variant. Both variants show an operational cost premium for FCEBs under current circumstances, with California highlighted on the graph for reference [23,24] . However, as hydrogen production increases, costs will go down. If costs reach the levels predicted in [6] FCEB operational costs could reach parity to BEBs. Note that the operational cost premium of the double-deck FCEB is less sensitive to the price of hydrogen than the single-deck.

Figure 4. Double-deck FCEB vs BEB: operational cost delta per passenger-

km as a function of hydrogen/electricity price.

Figure 5. Single-deck FCEB vs BEB: operational cost delta per passenger-km

as a function of hydrogen/electricity price.

Range & Passenger Carrying Capability

As in the previous section, the consumption values in Table 6 are

converted to a maximum range for each vehicle and drive cycle and plotted in Figure 6.

Figure 6. Predicted vehicle range on four real bus routes.

Page 6 of 11

01/11/2016

As expected, for both vehicles the single-deck variant has a larger average range than the double-deck (21 % for FCEB and 5 % for BEB) due to the lesser mass of the single-deck.

Whilst carrying the same number of passengers as the BEB, the FCEB on average has a 90 % larger range for the single-deck, and 65

% larger for double-deck.

The FCEB and BEB expected range follows the same trend for the first three drive cycles but diverge for the Rancagua route. Despite this route being an intensive hill ascent, it is the route with the largest range for the BEB whilst being second to worst case for the FCEB. This is a result of the much larger battery in the BEB which can accept larger currents than the FCEB. Thus, the regenerative braking energy recuperation potential is higher and the BEB can take better

advantage of the steep descent on the return journey of the drive cycle. The FCEB must activate the brake resistor or use a greater proportion of friction brakes to avoid overloading the battery. Although BEBs have a smaller energy storage on board for the ascent, in severe ascent/descent drive cycles this can be compensated for with an increase in efficiency from regenerative braking.

Although the single-deck variants have a greater absolute range, this trend is again reversed when factoring in the number of passengers.

Looking at the range in passenger-km (Figure 7), the single-deck performs worse (13 % for FCEB and 25 % for BEB). Thus, despite the double-deck vehicles having a smaller absolute range, they are more range efficient considering passenger-km. The trend is less pronounced for the FCEB as the single-deck variant has increased hydrogen storage capacity.

Figure 7. Predicted vehicle range on four real bus routes, normalised by

passenger loading.

Level of Service

Ideally for operators, new clean vehicle technologies would be able

to replace their existing fleet of diesel vehicles without any change to the level of service provided or fleet size required. Due to the range and infrastructure restrictions, this is unlikely to be the case in most

circumstances. For three of the four real bus routes logged for this work, the frequency of operation per bus is known. The buses used were diesel hybrids, and thus the frequency of their operation will be representative of normal levels of service for that route. Service information is listed in Table 7.

Table 7. Total time and distance covered per route each day.

Trips per day Total Time Total

Distance (km)

London 1 7 23:02:30 215

London 2 7 13:59:04 173

Santiago 6 17:36:00 300

Figure 8 shows how the modelled vehicles compare to the current level of service, in terms of the number of full trips they are capable of before running out of charge or fuel. On all three routes, both variants of the FCEB can comfortably provide the same level of

service as with the existing diesel fleet with no additional vehicles needed. The quick refuelling time of around 10 minutes [7] is useful for quick turnaround times on routes with a high degree of vehicle utilisation.

Figure 8. Number of trips possible for each vehicle on three real bus routes,

compared to the number of trips required for parity with existing service

levels.

The BEB is only capable of meeting the required range on the “London 2” cycle, indicating that in terms of maintaining service levels, BEBs are best suited to lighter duty/shorter routes. It is close to providing parity of service on “London 1”, however factoring in

the distance to the bus depot, future battery degradation and a factor of safety it is unable to provide the 7 trips without topping up the

Page 7 of 11

01/11/2016

battery. Additionally, the bus used for the “London 1” route has a very high degree of utilisation, not leaving enough time for overnight charging. Based on a depot charging rate of 100 kW [25], a full charge of the BEBs considered in this study would take 2 hours and 22 minutes leaving it unsuitable for routes with such high utilisation.

These results are in line with the International Renewable Energy Agency who state that fuel cells are most suited to heavy duty vehicles, long distances and high utilisation, while the opposite is true for the BEB [26]. Many larger buses fall in to both categories by having shorter distances than most heavy-duty applications, but potentially higher utilisation.

The large range and low refuelling times of FCEBs result in the most attractive option for replacing a diesel fleet with the goal of

minimising disruption to current service levels. In many cases, to maintain service levels with a fleet of overnight charging BEBs will necessitate a surplus of vehicles at an increased capital expenditure. A full total cost of ownership assessment is required to determine if the additional BEBs required outweighs the increased operating and capital costs of FCEBs.

The introduction of either FCEBs or BEBs as a local zero emission option will require the addition of supporting infrastructure. The cost

and availability in providing this will also factor into the individual operator’s decision on which technology to adopt. Opportunity charging BEBs offer another alternative, however these have major infrastructure modification requirements [7].

Environmental Perspective

Reducing greenhouse gas (GHG) emissions and improving local air quality are key motivations behind the uptake of alternative powertrain technologies [27]. Operators seeking to identify suitable low emission powertrains often compare vehicle emissions using life cycle assessment (LCA) methods. Well-to-wheels (WTW) analysis is

seen as the dominant LCA method for comparing alternative vehicle technologies and is widely used for policy support in road transport [28]. WTW analysis focuses on the life cycle processes of the energy carrier (e.g. diesel, hydrogen, electricity) used to propel the vehicle during operation. The WTW phase can be further broken down into the well-to-tank (WTT) and tank-to-wheel (TTW) phases. The TTW stage considers the on-board energy conversion to drive the vehicle based on the lifetime distance travelled, fuel energy required and vehicle efficiency [29]. The vehicles assessed in this paper produce

zero TTW air quality or GHG emissions.

The WTT stage includes the recovery or production of the feedstock for the energy carrier and subsequent energy conversion, delivery, transmission and storage. The production of the energy carrier as a transport fuel will require a primary energy input, either from a fossil fuel feedstock or a renewable energy source, and result in the direct or indirect production of GHG emissions. Vehicles powered by electricity or hydrogen produced via fossil fuel sources could have

the unintended consequence of producing higher emissions from a WTW perspective than internal combustion engine counterparts [30].

Hydrogen is typically produced from one of two methods: steam methane reforming (SMR) and the electrolysis of water (WE). 48% of global hydrogen sources are produced via SMR, WE accounts for 4% and the rest is a result of fossil oil and coal gasification [26]. For WE, the energy demand required to split water is high and it is therefore imperative to use electricity generated from renewable

sources to produce low WTW emissions.

GHG emissions per passenger-km for the real-world UK and Chile based drive cycles have been modelled using a WTW approach, with GHG emissions are quantified as a carbon dioxide equivalent value, CO2e (Figures 9 & 10). The impact that the CO2e intensity of the electricity supply has on both WTT hydrogen emissions from WE

and subsequent GHG emissions per passenger-km is demonstrated. Assumptions used in this calculation are detailed in the Appendix. In the interest of a fair comparison, the WTW impacts of BEB technologies are also displayed. Both Figures 9 & 10 show that from the WTW perspective, FCEBs have a higher environmental burden than their BEB counterparts if hydrogen produced by WE comes from the same electricity grid that charges the BEBs. On a per passenger-km basis, GHGs from double deck vehicles are lower than

their single deck counterparts. Notably, hydrogen produced via SMR has a lower GHGs than hydrogen produced via WE using each countries’ current electricity grid.

These results are intended to illustrate various environmental impact ranges of low emission technologies and does not mean hydrogen as a transport fuel should be discarded. For instance, to create a new export commodity, Chile is pursuing a clean hydrogen economy. Chile aims to leverage abundant solar electricity resources to produce

low cost hydrogen, enabling the transition to lower carbon fuels and diversifying the economy [6]. Electricity generated via sustainable or renewable sources combined with zero tailpipe emission technologies is seen as a clear pathway for reducing the environmental burden of transport and increasing energy security [30]. For locations with an unreliable or high emission factor grid electricity supply, the prospect of importing clean hydrogen for an FCEB fleet offers the opportunity of reliable low overall emissions where BEBs could not.

Figure 9. WTW GHG emissions sensitivities of UK based cycles.

0

10

20

30

40

50

60

70

80

90

100

0 200 400 600 800 1000WT

W G

HG

em

issi

on

s p

er p

ass

enger

-km

(gC

O2e/

pk

m)

Electricity grid emission factor (gCO2e/kWh)London #1-BEB-SD London #1-BEB-DDLondon #1-FCEB-SD London #1-FCEB-DDLondon #2-BEB-SD London #2-BEB-DDLondon #2-FCEB-SD London #2-FCEB-DD

02000

0 9814 19628 29442 39256 49070Axis

Tit

le H2 via WE emission factor (gCO2e/kgH2)

UK

gri

d C

O2e

inte

nsi

ty

Win

d e

lect

rici

ty C

O2e

inte

nsi

ty

H2 v

ia S

MR

Page 8 of 11

01/11/2016

Figure 10. WTW GHG emissions sensitivities of Chile based cycles.

Summary

Vehicle models for a future production Wrightbus FCEB and BEB were developed for the prediction of fuel/energy consumption. Hydrogen consumption was validated based on a prototype FCEB vehicle. Predicted fuel/energy usage and range of the vehicles was shown for four industry standard drive cycles to act as a performance

benchmark.

Four drive cycles were created through logging of real bus routes then simulated with the vehicle models. The following averaged findings were noted based on the modelled outputs (note that all figures given are predicted values from modelling efforts and may differ from actual fuel/energy consumption.):

1. Single-deck variants used around 5 % less fuel/energy than their double-deck counterparts. However, when expressed on a per-

passenger basis, this trend is reversed. The FCEB single-deck is on average 32 % less efficient per passenger-km then the double-deck. Similarly, the BEB single-deck is 33 % less efficient than the double-deck.

2. The operating cost of the FCEB was greater than the BEB in current circumstances. However, as hydrogen production increases, costs will go down. If costs reach the levels predicted in [6] FCEB operational costs could reach parity to BEBs. The

operating cost of the single-deck FCEB is more sensitive to the price of hydrogen than the double-deck.

3. The single-deck variants had a larger average range than the double-deck (21 % for FCEB and 5 % for BEB). Despite this, the double-decks are more range efficient. Single-decks performs worse per passenger-km (13 % for FCEB and 25 % for BEB).

4. Whilst carrying the same number of passengers as the BEB, the FCEB had a 90 % larger range for the single-deck, and 65 % larger for double-deck.

5. For the routes considered, only the FCEB had the required range to maintain current levels of service even on longer routes with

higher vehicle utilisation. The BEB was capable on one route indicating suitability for light duty/shorter routes. These findings agree with the International Renewable Energy Agency [26].

A WTW analysis based on the modelled outputs for the four real bus routes (from Chile and the UK) showed:

1. On a per passenger-km basis, GHGs from double deck vehicles are lower than their single deck counterparts

2. FCEBs have a higher environmental burden than their BEB

counterparts if hydrogen produced by water electrolysis comes from the same electricity grid that charges the BEBs. For locations with a high emission factor grid, importation of hydrogen from a renewable source can mitigate this.

The optimum local zero emission bus depends on many factors and is specific to each situation. Consideration should be taken to whether current levels of service can be met, and whether the additional capital cost of a FCEB and current higher operating costs outweigh

the potentially larger fleet of BEBs required for parity of service. Environmental impacts are highly sensitive to geographical location and the emission factor of hydrogen/grid electricity supply. A future work recommendation to help quantify the above considerations is a full life cycle assessment including total cost of ownership, considering the modelled outputs discussed here.

References

1. Hydrogen Europe, “Jive/Jive 2,” https://hydrogeneurope.eu/jivejive-2, accessed Sep. 2019.

2. Zero Emission Urban Bus System, “ZeEUS eBus Report #2,” https://zeeus.eu/publications, accessed Nov. 2019.

3. H2 Aberdeen, “Hydrogen Bus Project,” http://www.h2aberdeen.com/, accessed Sep. 2019.

4. Hebditch, J., The Press and Journal, “World first for Aberdeen

as city orders 15 double decker hydrogen buses,” July 2019, https://www.independent.co.uk/news/uk/home-news/hydrogen-bus-doubledecker-london-tfl-sadiq-khan-air-pollution-a8909326.html, accessed Oct. 2019.

5. Wyatt, T., Indpendent, “World’s first hydrogen double-decker buses coming to London’s streets,” May 2019, https://www.independent.co.uk/news/uk/home-news/hydrogen-bus-doubledecker-london-tfl-sadiq-khan-air-pollution-

a8909326.html, accessed Oct. 2019.

6. IRENA, “Hydrogen: A Renewable Energy Perspective,” International Renewable Energy Agency, Abu Dhabi, 2019, ISBN 978-92-9260-151-5.

7. Mahmoud, M., Garnett, R., Ferguson, M., and Kanaroglou, P., “Electric buses: A review of alternative powertrains,” Renewable and Sustainable Energy Reviews 62:673–684, 2016, doi:10.1016/j.rser.2016.05.019.

0

10

20

30

40

50

60

70

80

90

100

0 200 400 600 800 1000WT

W G

HG

em

issi

on

s p

er p

ass

enger

-km

(gC

O2e/

pk

m)

Electricity grid emission factor (gCO2e/kWh)Santiago-BEB-SD Santiago-BEB-DDSantiago-FCEB-SD Santiago-FCEB-DDRancagua-BEB-SD Rancagua-BEB-DDRancagua-FCEB-SD Rancagua-FCEB-DD

02000

0 9814 19628 29442 39256 49070Axis

Tit

le H2 via WE emission factor (gCO2e/kgH2)

Sola

r P

V C

O2e

inte

nsi

ty

Chil

e S

IC g

rid C

O2e

inte

nsi

ty

H2 v

ia S

MR

Page 9 of 11

01/11/2016

8. Correa, G., Muñoz, P.M., and Rodriguez, C.R., “A comparative energy and environmental analysis of a diesel, hybrid, hydrogen and electric urban bus,” Energy 187:115906, 2019, doi:10.1016/j.energy.2019.115906.

9. Chang, C.-C., Liao, Y.-T., and Chang, Y.-W., “Life cycle

assessment of alternative energy types – including hydrogen – for public city buses in Taiwan,” International Journal of Hydrogen Energy 44(33):18472–18482, 2019, doi:10.1016/j.ijhydene.2019.05.073.

10. Lee, D.-Y., Elgowainy, A., and Vijayagopal, R., “Well-to-wheel environmental implications of fuel economy targets for hydrogen fuel cell electric buses in the United States,” Energy Policy 128:565–583, 2019, doi:10.1016/j.enpol.2019.01.021.

11. Stevens, G., Murtagh, M., Kee, R., Early, J. et al., “Development of a Vehicle Model Architecture to Improve Modeling Flexibility,” SAE Int. J. Engines 10(3):1328-1366, 2017, https://doi.org/10.4271/2017-01-1138.

12. Murtagh, M., Early, J., Stevens, G., Cunningham, G. et al., “Modelling and Control of a Hybrid Urban Bus,” SAE Technical Paper 2019-01-0354, 2019, https://doi.org/10.4271/2019-01-0354.

13. Stevens, G., Early, J., Cunningham, G., Murtagh, M. et al., “Multi-fidelity validation algorithm for next generation hybrid-electric vehicle system design:,” Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering 233(13), 3438–3448, 2019, doi:10.1177/0954407018825015.

14. Zhang, Y., Murtagh, M., Steele, D., Cunningham, G. et al., “Optimal Control Strategy for a Next Generation Range

Extended Electric Bus,” SAE Technical Paper 2020-01-0844, 2020, https://doi.org/10.4271/2020-01-0844.

15. Blades, L., Douglas, R., Early, J., Lo, C.Y., and Best, R., “Advanced Driver-Assistance Systems for City Bus Applications.,” SAE Technical Paper 2020-01–1208, 2020, https://doi.org/10.4271/2020-01-1208.

16. Cole, D., “StreetAir”, Bus & Coach Buyer, Aug. 2016, https://www.busandcoachbuyer.com/streetair/, accessed Oct. 2019.

17. “London unveils first fuel cell double-decker bus, from Wrightbus”, Fuel Cells Bulletin 2016(12):3, 2016, doi:10.1016/S1464-2859(16)30337-6.

18. Ballard, “FCMove,” https://www.ballard.com/fuel-cell-solutions/fuel-cell-power-products/fcmove, accessed Oct. 2019.

19. LowCVP, “LEB Certificates,” https://www.lowcvp.org.uk/Hubs/leb/LEBCertificates.htm, accessed Oct. 2019.

20. LowCVP, “Testing & Accreditation Procedures,” https://www.lowcvp.org.uk/Hubs/leb/TestingandAccreditation/TestingAccreditationProcedures.htm, accessed Oct. 2019.

21. UITP, “SORT BROCHURE,” https://www.uitp.org/sort-brochure, accessed Oct. 2019.

22. Morrison, G.M. and Rauch, S., “Highway and Urban Environment: Proceedings of the 8th Highway and Urban Environment Symposium,” (Springer Science & Business

Media, 2007), ISBN 978-1-4020-6010-6.

23. California Fuel Cell Partnership, “Cost to Refill.” https://cafcp.org/content/cost-refill, accessed Nov. 2019.

24. CalChamber ALERT, “California Electricity Prices: No. 7 in U.S,” CalChamber Alert, accessed Nov. 2019.

25. Göhlich, D., Fay, T.-A., Jefferies, D., Lauth, E. et al., “Design of urban electric bus systems,” Design Science 4(15), 2018, doi:10.1017/dsj.2018.10.

26. IRENA, “Hydrogen From Renewable Power, Technology Outlook for the Energy Transition,” Abu Dhabi, 2018, ISBN 978-92-9260-077-8.

27. LowCVP, “The Low Emission Bus Guide,” (London, Low Carbon Vehicle Partnership, London, 2016).

28. Moro, A. and Helmers, E., “A new hybrid method for reducing the gap between WTW and LCA in the carbon footprint assessment of electric vehicles,” International Journal of Life

Cycle Assessment 22(1):4–14, 2017, doi:10.1007/s11367-015-0954-z.

29. MacLean, H.L. and Lave, L.B., “Evaluating automobile fuel/propulsions system technologies,” Progress in Energy and Combustion Science 29(1):1–69, 2003, doi:10.1016/S0360-1285(02)00032-1.

30. Nordelöf, A., Messagie, M., Tillman, A.M., Ljunggren Söderman, M., and Mierlo, J. Van, “Environmental impacts of

hybrid, plug-in hybrid, and battery electric vehicles - what can we learn from life cycle assessment?,” International Journal of Life Cycle Assessment 19(11):1866–1890, 2014, doi:10.1007/s11367-014-0788-0.

31. Edwards, R., Larivé, J.-F., Rickeard, D., and Weindorf, W., “WELL-TO-TANK Report Version 4.a. JEC well-to-wheels analysis,” JRC Technical Reports, 2014, doi:10.2790/95629.

32. Armijo, J. and Philibert, C., “Flexible production of green hydrogen and ammonia from variable solar and wind energy.

Case study of Chile and Argentina,” 2019, doi:10.13140/RG.2.2.36547.66081.

33. Department for Business Energy and Industrial Strategy, (BEIS), “Greenhouse gas reporting: conversion factors 2019,” https://www.gov.uk/government/publications/greenhouse-gas-reporting-conversion-factors-2019, accessed Nov. 2019.

34. Energía Abierta, “Factor de Emisión - Promedio Mensual,” http://energiaabierta.cl/sustentabilidad/?lang=en, accessed Nov.

2019.

35. Schlomer, S., Bruckner, T., Fulton, L., Hertwich, E. et al., “Annex III: Technology-Specific Cost and Performance

Page 10 of 11

01/11/2016

Parameters,” Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change 1329–1356, 2014, doi:https://doi.org/10.1017/CBO9781107415416.025

Contact Information

Sir William Wright Technology Centre, 50 Malone Road, Queen’s University Belfast, Belfast, BT9 5BS, United Kingdom

email: wtech@qub.ac.uk

Acknowledgements

The authors would like to thank and acknowledge Wrights Group Ltd., as well as the Engineering and Physical Sciences Research Council Grant EP/S036695/1, for funding this research.

Definitions/Abbreviations

BEB Battery electric bus

CO2e Carbon dioxide equivalent

FCEB Fuel cell electric bus

GHG Greenhouse gas

HVAC Heating, ventilation and

cooling

LCA Life cycle assessment

LEB Low Emission Bus

LTO Lithium-titanate

NMC Lithium-nickel-manganese-cobalt-oxide

PV Photovoltaic

RPA Relative positive acceleration

SIC Sistema Interconectado Central

SMR Steam methane reforming

SORT Standardised on road test

TTW Tank-to-wheels

WE Water electrolysis

WTT Well-to-tank

WTW Well-to-wheels

Page 11 of 11

01/11/2016

Appendix

For the WTW emission calculations, the following assumptions are used:

• Electricity demand per kg hydrogen produced via water electrolysis is calculated as 49.07 kWh/kg. This assumes a lower heating value of hydrogen of 120.1 MJ/kg [31], an electrolysis process efficiency of 70% [32] and a compression efficiency of 97% [31]. Values are

assumed the same regardless of geographical location.

• Emission factor of hydrogen produced via steam methane reforming is based on average EU conditions [31], calculated as 13.8 kgCO2e/kgH2.

• UK grid emission factor calculated as 316 gCO2e/kWh [33]. Chilean Sistema Interconectado Central (SIC) grid emission factor is taken as 336 gCO2e/kWh [34]. The SIC grid provides electricity to the central part of the country, including Santiago and Rancagua. Both emission

factors are based on latest publicly available values.

• Solar photovoltaic (PV) and wind based electricity emission factors are median life cycle GHG emissions values reported by selected electricity supply technologies by the Intergovernmental Panel on Climate Change [35]. Values are assumed the same regardless of

geographical location.