April 10, 2022 by Steve

TTC eBus Study: Final Results

Updated April 10, 2022 at 10:30pm: Minor typos fixed. Bus order size for eBuses by TTC corrected. Reference to use of pantographs for charging on TTC buses corrected.

Updated April 18, 2022 at 2:50 pm: Note that the GHG savings cited in the TTC’s chart below are off by a factor of 1,000 because they mixed up kilos and tonnes partway through the calculation. See also TTC eBus Errata: Tonnes and Kilos Are Different.

Since June 2019, the TTC ran head-to-head trials of three manufacturers’ battery-electric buses with a fleet of 60 vehicles:

New Flyer models SR2304 (10) and SR2382 (15)
Proterra models Catalyst 40 E2 RR Pro Drive (10) and DuoPower (15)
BYD model K9M (10)

Nova Bus was not part of the trial because, when it was launched, they did not have a vehicle with sufficient range to meet the specifications. However, their hybrid diesel-electric bus, of which the TTC has many, was used as a comparator for the trial.

The low number of BYD buses was due to their inability to supply vehicles even though their lobbyists had engineered, through Deputy Mayor Minnan-Wong, a “deputation” at a TTC Board Meeting that turned into a full sales pitch clearly hoping to short-circuit the procurement process. This was not a high point in TTC history, and the move to a green fleet was launched under very dubious circumstances.

There is also bitter irony for those who remember TTC history. Three decades ago, the TTC opted for the allegedly-green technology of Natural Gas buses as a replacement for trolley bus system expansion. The CNG buses are long gone from Toronto, and the TTC now plans to move completely to electric transit. Lobbyists are good at selling things – whether they work or not is a secondary consideration.

The 102 page report TTC’s Green Bus Program: Final Results of TTC’s Head-to-Head eBus Evaluation goes into great detail of Toronto’s experience with their trial fleet and sets out many “lessons learned” and “must haves” for any large-scale procurement. This article is organized somewhat like the report with an overview followed by some of the technical background. The “lessons learned” have been consolidated at the end. Interested readers should consult the full report.

The clearly superior vehicles in the trial were the New Flyer buses. There were severe problems with reliability and maintainability in both the Proterra and BYD fleets, and some of the “must haves” would exclude them from consideration even if both vehicles and manufacturers had performed better.

Whether this technical outcome will be coloured by another round of lobbying remains to be seen. There will be a lot of money sloshing around as governments rush to “buy green”, but running transit requires a fleet that delivers reliable service, not just publicity photos. Toronto cannot afford to tie the future of its bus fleet to a manufacturer whose political connections outweigh their ability to deliver good products.

The TTC has funding in place from various governments to cover the purchase of about 600 vehicles. These will meet its replacement and growth needs from 2022 to 2025. In February 2022, the TTC ordered 336 buses for delivery by the end of 2023:

Nova Bus LFS Hybrid 40′ (134)
New Flyer Xcelsior Hybrid 40′ (134)
New Flyer Xcelsior Hybrid 60′ (68)

These will be the last buses with diesel propulsion for Toronto, and they would be due for replacement in the mid 2030s completing the conversion to an all-electric bus fleet.

An RFP (Request for Proposals) was issued on April 4, 2022 for a large purchase (at least 240 vehicles) of eBuses with contract award planned for the third quarter of 2022. This lands in the middle of the municipal election campaign, and the authority to award will be delegated to TTC management by the Board. Bids will close on June 17, and the successful vendor(s) would be notified in July with execution of agreements in August. (The last scheduled TTC Board meeting is on July 14, 2022.)

The specification for these buses was developed jointly by the TTC with other agencies:

The TTC is engaged with other peer transit agencies in the province, including Brampton Transit, Mississauga Transit, York Region Transit, and others through the Ontario Public Transit Association on the first interagency co-operative procurement of eBuses. The aim of this collaboration is to develop a single zero-emissions bus procurement specification with the immediate benefit of reducing cost through economies of scale. The long-term benefit is through the optimization and standardization of customer experience and, operations and maintenance throughout the GTHA and beyond.
TTC Report at p. 4

The potential quantity of buses is considerably higher with options for both the TTC and other agencies.

Source: TTC Vendor Briefing Session Deck April 8, 2022 at p. 5

In parallel to its migration to an electric fleet, the TTC must convert its bus garages including the provision of charging infrastructure for hundreds of vehicles at each location. At a previous meeting, the Board authorized an agreement with Ontario Power Generation and Toronto Hydro for the charging infrastructure. The utilities will build, own and maintain this as an extension of their distribution system.

Although the specification includes a requirement for on-route charging using stationary charging points, the TTC has not yet determined if or how such facilities would be used. There is no consideration of “in motion” charging using conventional trolleybus infrastructure to avoid the need for buses to lay over to recharge during their revenue service hours.

The overall plan for both buses and charging infrastructure is shown in the table below grouped by garage. This accounts for 1,085 buses, about half of the existing fleet. The first 240 buses planned in the contract award this year would take the TTC into early 2025. The program is not funded yet beyond that point. As and when more money appears, the TTC would extend its order.

Note that there are two phases to the installation of charging facilities at garages as the roll out of electrification works its way through the system. This allows some routes from each garage to operate with eBuses earlier in the program than might be practical if the conversion went garage-by-garage over the next decade.

The two-step scheme would also allow for a tactical change in charging strategy to move more of this to enroute facilities such charging stations at terminals. Although the TTC report is mostly silent on any charging technique beyond garage-based plug-in systems, there is a reference in “lessons learned” to a conversion to pantograph charging as a cure for problems with charging cables.

The TTC projects that life cycle costs for an electric fleet will be lower than for the diesels and hybrids it will replace because both energy and maintenance costs will go down. By 2040 this would save about $167 million annually. These estimates are sensitive to the future price of diesel fuel compared to electricity, but the TTC has not shown a range of values to indicate what the effect might be.

An 18-Year Design Life

Although the TTC report does not mention this, the actual RFP includes an interesting specification for fleet longevity. This signals a return to 18-year lifespans for the bus fleet after a retreat to 12 years in current fleet planning. If this can be achieved, it will offset the higher capital cost of the vehicles compared to hybrids or diesel buses.

1.1.1 The Bus shall have an 18-year design life and be equipped with a long life structure in accordance with Specification Section 1.8, made from full stainless steel in accordance with Specification Section 3.0, have a body with a maximum overall length of 12.8 m (42ft.) including a stowed Bike Rack , 2.59 m (8 ft.-6 in.) in width and a maximum overall height of 3.4 m (134 in.).
RFP Technical Requirements Section 1, Page 6

Later in the RFP:

1.8 SERVICE LIFE
Buses shall be designed for a minimum service life of 18 years or 1,610,000 km (1,000,000 mi.), under severe operating conditions similar to revenue transit operation in the City of Toronto.
RFP Technical Requirements Section 1, Page 19

And in more detail:

The vehicle design life shall be validated by successful completion of a simulated 12-year average New York City duty cycle service life. The test program shall be designed around input measurements taken from a vehicle configured similarly to the test vehicle while it’s being operated over a known severe route. The New York City B.35 route or an approved equivalent (i.e., the Queens Q.44 route is now reportedly used by New York City), shall be used for a simulated 800,000 km (500,000 mi.) to demonstrate Bus longevity. This is generally considered to be the equivalent of 16 to 18 years operating life at all
other transit properties.
RFP Technical Requirements Section 1, Page 20

“Must Have” Specifications

An important outcome of the trial has been the development of a “must have” list, and certain aspects of any new fleet are not negotiable. Toronto has a history with every type of vehicle (subway, streetcar, bus) where pervasive problems have hobbled fleet performance and availability.

TTC’s next large-scale eBus procurement includes ‘must have’ requirements that are informed by the head-to-head evaluation and focus on ensuring longevity of the bus structure and high system reliability through a proven platform (e.g. stainless steel structure, doors, HVAC, suspension, etc.).
TTC Report at p. 2

These requirements are:

1. Altoona and shaker table testing has been successfully completed;
2. A full stainless steel structure with a minimum of six years of in service experience;
3. A minimum usable battery capacity of 400 kWh;
4. A maximum overall bus length of 12.8 m (42 ft.) including a stowed bike rack;
5. A maximum overall height of 340 cm (134 in.) including any roof-mounted equipment;
6. Ability to charge via roof mounted pantograph charging interface, capable of accepting a minimum charge rate of 300kW (400 ADC) at 750 VDC or greater via SAE J3105/1; and
7. Two rear-mounted charging ports capable of accepting a minimum charging rate of 150 kW (200 ADC) at 750 VDC or greater via SAE J1772.
TTC Report at p. 24

Requirement 3 conflicts with statements elsewhere in the report where a maximum length of 40 feet is cited so that buses will fit within existing garage designs and operations. The difference appears to be in whether the bike rack counts toward the total, but it is not clear whether the Proterra bus would meet this requirement.

Physical Compatibility: The industry standard bus length is 40-feet (12 metres). This standard was used to design storage facilities in the TTC’s existing bus garages.
The Nova HEV, BYD, and NFI buses meet this standard. Proterra buses are 42.5 feet long, but also offers the highest seating and standee capacity. Based on our bus garage layout, procurement of additional Proterra buses would result in a loss of storage capacity of approximately 10% at four of eight garages. The remaining four bus garages could accommodate this additional length. However, this would impose a significant operational constraint that would prevent movement of buses between garages.
TTC Report at p. 15

A maximum bus length specification of 40 feet is required in order to preserve bus storage density at existing maintenance facilities; …
TTC Report at p. 16, also p. 28, under “Lessons Learned”

It is not clear whether the TTC is prepared to accept buses over 40 feet long, and what position they will take about Proterra vehicles on that account. Other issues with that vendor, notably bus reliability, might knock them out of the running regardless of bus length.

An additional requirement applies to the contract itself rather than to the buses, and it addresses the City’s equity goals:

In support of the commitment to diversity, equity, and inclusion, the Contractor must agree, as a fundamental component to the Contract, to meet the Procurement Equity Requirements, by applying a percentage of the Contract Price in respect of the Diverse Business Enterprise Requirement and a specified number and percentage, as stated in the Proposal, in respect of the Equity Hired Requirement.
TTC Report at p. 7

This is a contrast to recent provincial actions to back away from equity and community benefit components in contracts.

Technical Review

Vehicle Reliability

Of particular interest in the review of three electric fleets, the majority of problems lay not with the propulsion and charging systems, but with the physical bus and subsystems that are common to all vehicles such as doors. This suggest that the primary skill for a vendor is simply being able to build a vehicle that does not fail physically, as opposed to one with whatever the latest technology might be.

The mean distance to failure typically starts out low with new vehicles thanks to problems that only appear in revenue service, but then should build up to better results. Note that a “failure” is defined as a fault that requires a bus to go out of service and/or causes a service delay of five minutes or more.

The chart below compares performance of the NOVA Hybrid Electric buses now in service with the three electric fleets. Note that data for each fleet are constrained by the point when vehicles were first available for service with BYD bringing up the rear. Flyer vehicle stats continue to improve while BYD and Proterra are on a downward trend. None has reached the level attained by the Hybrids, although Flyer is headed into that territory.

The RFP for new eBuses specifies a target of 30,000 km MDBF. The higher the number actually achieved, the less time buses spend in the garage for repairs and the fewer vehicles are required to provide maintenance spares.

“Availability” is defined as a bus being available for service rather than being offline for manufacturer repairs/fixes under warranty. BYD’s stats were particularly affected by long lead times for supply of replacement parts, and some buses were out of service for extended periods.

Note that an “available” bus might still not be in revenue service because it is in a routine maintenance cycle. This metric reflects vendor behaviour and vehicle reliability during the acceptance and warranty period.

A need for repair, whether it be counted towards a service disruption as above, or simply as a problem noticed in the garage, is not simply an incident, but a period of time a bus is out of service. The longer this time, the greater the overhead of carrying spare buses, and the added need for garage space while a vehicle is repaired. The lack of available spare parts was a big issue on this account.

Vehicle Range and Energy Consumption

An essential component of fleet and operational planning is the range of a vehicle. This is not an issue for subways and streetcars which do not have an onboard energy source and could run indefinitely. Buses (except for trolley buses that have not been used in Toronto for three decades) must be refuelled either for diesel or for electricity before they exhaust onboard capacity. The full theoretical capacity cannot be assumed for service planning because consumption rates differ from route to route, are affected by traffic conditions and weather, and can even be influenced by an operator’s driving behaviour. All of this is well-known to anyone familiar with automotive fuel consumption.

A further issue lies in energy demands beyond simply moving the vehicle, notably the operation of heating and cooling both for passengers and for onboard equipment. This is a substantial additional draw particularly in winter.

If schedules must take into account the need to refuel midway through a shift, this adds to complexity and to the non-productive hours for vehicles and their crews. The situation is different in Toronto where a large proportion of the bus fleet remains in service all day compared to a system which primarily exists for peak demand and sees most buses back at the garage where they can be “topped up” if need be.

(In January 2020 on pre-pandemic schedules, mid-day and early evening bus requirements were about 2/3 of the peak period levels. In March 2022, the ratio is even higher.)

This is generally not a problem for the diesel or diesel/hybrid fleets except, possibly, for buses that stay out for 24 hours from a morning peak through night service, and it can be avoided by scheduling. The situation for electric buses, with a smaller and variable range is quite different.

There is a need for TTC operations to accurately determine the expected range of the buses. This is difficult given how many factors influence the efficiency and range of an eBus. The TTC attempted to estimate the range based on the in-service data through correlating the State of Charge (SoC) used per km and extrapolating to the full SoC range available. It must be noted that this is an oversimplification to estimate the range and SoC is a calculated value from the OEMs that we have found to not be entirely accurate. However, this simple calculation can be used to highlight the variation in range due to seasonal changes as well as to the large variation in efficiency as a result of the factors mentioned above.
TTC Report at pp 52-53

Energy consumption varies both by bus model and by season. The charts below show the distribution of values for effective battery capacity (the energy that can actually be used on one charge) and the service range based on battery capacity and energy consumption characteristics. (Click on a chart to open each series in a gallery to step between each seasonal chart.) From a planning point of view, a consistent range is preferable across seasons so that the TTC does not need separate “winter” and “summer” schedules.

The data are summarized in the table below the charts. Note that these values are well below manufacturers’ claims for their products.

Energy consumption is affected by route characteristics like speed, stop spacing, grades and passenger loads. The available range coupled with vehicle reliability can lead to buses having different operational stats depending on the routes that they serve. These factors must be teased apart to isolate geographic factors from inherent performance of the vehicles. (A similar problem has existed for years in comparisons of bus vs streetcar and trolleybus operations that tended to overstate relative electric vehicle costs because of the routes on which they were used.)

The Proterra fleet was based at Mount Dennis Garage which disproportionately serves inner city routes with lower operating speeds compared to Arrow Road (New Flyer) and Eglinton Garages (BYD). Although the Proterra buses ran comparable or better hours/day than the other vendors’ buses, they accumulated less mileage because they operated at a lower average speed.

Energy consumption was affected by the routes each fleet operated on and also by the seasonal heating and cooling loads. When the powertrain energy is broken out from the total, clear peaks due to vehicle heating in winter months are obvious. This is not an issue for New Flyer because it uses an auxiliary diesel generator for heating power rather than drawing on the batteries. This engine is much smaller than those used for propulsion in diesel or diesel-hybrid buses, and it is not subject to the stress of acceleration cycles.

The energy consumption is sensitive to the ambient temperature for BYD and Proterra buses, but not for New Flyer. This contributes to a more reliable vehicle range over the seasons. This is obviously a larger issue in cities with highly variable climates such as Toronto.

The effect of the different approach to heating in the New Flyer eBuses is clear in the chart of diesel consumption during the winter season. Diesel consumption was higher because this subsystem provided vehicle heating, but the offsetting benefit was that battery capacity was more predictable.

One note about this, however, is that the vehicle specification calls only for enough diesel capacity to operate the heating system for eight hours. This could compromise the range of a bus during cold weather even if the battery had sufficient capacity.

Like all modern electric vehicles, the eBuses use power regeneration during braking to recapture kinetic energy from the vehicle. This has been common on subway trains and streetcars for decades, and it is not a new technology. The regenerated power shows a seasonal variation especially for Proterra and BYD because in the winter, that power goes into the heating system rather than back into the batteries.

Finally, there is some interaction between precipitation and energy consumption, although this is likely due to driving conditions during periods of heavy precipitation. Nonetheless, it is a consideration that during periods of very bad weather, vehicle range could be lower than normal.

Charging Systems

The chargers used for New Flyer and Proterra eBuses at Arrow and Mount Dennis Garages were ABB DC Fast Chargers. At Eglinton Garage, BYD AC chargers were used, although this will not be the case for any future orders that BYD might supply.

At all three sites chargers had an availability above 98%. The majority of failures were caused by cables either being run over by buses (ABB) or small wires breaking from repeated flexing (BYD, since replaced with better wiring). A variety of other problems are detailed in the report.

Of note is a remark under “lessons learned” that:

Cable management needs to be improved and included in future charging cables. The switch to pantograph charging will alleviate this issue for the eBus fleet.
TTC Report at p. 82

The use of “will” implies that the TTC contemplates a move from cable based to pantograph based charging at garages. This begs the question of whether this will occur as early as possible in the provision of new charging infrastructure and, indeed, if cable-based charging will disappear. This would simplify garage operations and consolidate charging technology for both garage and on route locations.

Passenger and Employee Response

The report includes sections on responses from both passengers and employees to the new buses. This text is edited and condensed from the TTC Report for brevity.

For passengers, the major points in bus design are:

Overall, 70% of riders are satisfied with current TTC bus design.
Almost two-thirds of riders prefer more seating capacity, flexible (flip-up) seating and forward-facing seating.
There is very strong support at 90% for seating capacity as an important factor.
Hygiene and comfort were ranked as important, although there are tradeoffs between fabric and plastic seats on these counts.
Over 60% of riders prefer a dedicated area for strollers and other large items.
More than half prefer more stanchions (vertical poles) to overhead handholds.
Almost half ranked digital screens as the most important technology compared to Wi-Fi at roughly one in six.
In an online survey, customers reported being satisfied with the three types of buses at different rates with New Flyer highest (91%), then BYD (83%) and Proterra close behind (79%).

For operators, the response to the eBuses varied

Over two thirds of the Mount Dennis operators were dissatisfied with the Proterra eBus when compared to a diesel bus. Operators at New Flyer and BYD garages were not strongly dissatisfied with those vehicles.
The BYD electric buses were the most favoured in the surveys completed, scoring high in seven categories: noise level, ergonomics, visibility/sightlines, ride comfort, acceleration, steering/manoeuverability and night driving. The only low score was for braking. The most common write-in comments were:
- Dissatisfaction with the braking systems;
- Glare/reflection on the front windshield during night driving; and
- The location of the cup holder.
The NFI electric buses scored high in two categories: noise level and braking. However, approximately 40% were dissatisfied with six categories: ergonomics, visibility/sightlines, ride comfort, acceleration, steering/manoeuverability and night driving. The most common write-in comments were:
- Braking and acceleration pedal locations, size and adjustment;
- Small size of the driver’s compartment; and
- Small size of the driver’ shield/barrier
The Proterra electric buses scored poorly in four categories: ergonomics, visibility/sightlines, ride comfort, steering/manoeuverability. They did score satisfactory in two categories: noise level and acceleration. The most common write-in comments were:
- Small size of the driver’s compartment;
- The turning radius of the bus; and
- The bus ramp kneeling issues.

For maintenance staff:

A high proportion of maintenance workers at Arrow (NFI) and Eglinton (BYD) were either positive or neutral about their fleets. At Mount Dennis (Proterra) about 15% were dissatisfied as compared to other buses at this garage.
At Eglinton (BYD), about 20% were dissatisfied with the following: diagnostic tools, maintenance manual content and navigation, parts manual content and navigation and layout of maintenance components. About 30% were satisfied with: noise level, visibility/sightlines, ride comfort, acceleration and steering/manoeuverability.
Arrow Road (NFI) had the most favourable responses scoring high in all the categories.
At Mount Dennis Garage (Proterra) high positive or neutral scores were found in four categories: noise levels, visibility/sightlines, ride comfort and acceleration, and for the remaining categories. The remaining categories also scored well, but not at quite as strong a level: steering/manoeuverability, diagnostic tools, maintenance manual content and navigation, parts manual content and navigation and layout of maintenance components.

Lessons Learned and Next Steps

Text in this section has been copied directly from the TTC report, but is consolidated here to pull all sections together. There are inconsistencies between the first instance of some of these points in the main report and the version in the technical appendix.

For example, the main report does not mention the maximum 42 foot bus length including the bike rack, but the appendix does. Many sections in the appendix are not carried over to the main report.

Corrosion Resistant Frame Structure (pp 16-17)

A maximum bus length specification of 40 feet is required in order to preserve bus storage density at existing maintenance facilities;
Bus specifications to require DC charging capability using SAE interface and communication standards to allow for maximum charge rates, competitive procurement, and interoperability between buses and chargers across all maintenance facilities; and
Stainless steel frame structure negates the need for, and associated risks of, annual rust-proofing maintenance programs.

Updates:

A maximum bus length specification of 42 feet including a stowed bike has been specified as part of the procurement pre-qualification criteria;
DC charging capability using SAE interface and communication standards has been specified as part of the procurement pre-qualification criteria; and
A stainless steel frame structure with a minimum of six years of in-service experience has been specified as part of the procurement pre-qualification criteria.

Distance Between Repairs

This section exists only in the appendix, not in the main report. It duplicates requirements that are under “Fleet Availability” below.

Fleet Availability (p 20)

Continue to monitor eBus availability performance, mature product with vendors and prioritize retrofit campaigns that will yield reliability and availability improvements.
BYD to hire a second field service technician in Q2 2021.
Include availability metrics to be achieved by the eBus OEM in future procurement contracts. Failure to meet the availability targets will result in liquidated damages.

Updates:

Second field service technician for BYD fleet support started working in Q3-2021; however, fleet availability continues to trend downward.
A minimum fleet availability target with associated liquidated damages has been incorporated into the next bus procurement contracts.

Duration to Final Acceptance (p 21)

Through a comprehensive review of commercial terms against industry peers and across modes (i.e. bus, subway and streetcar), the TTC is restructuring its milestone payments. Included in this restructure is a higher milestone payment percentage due at FAC in order to motivate vendors to improve quality and responsiveness during the acceptance process.

Updates:

The TTC has restructured its milestone payments for the next bus procurements as an approach to provide greater incentive for successful/on-time issuance of Final Acceptance Certificates. For the hybrid-electric bus procurement, TTC has moved away from a high percentage due upon delivery (75%) to the following:
i. From 0% to 20% upon Contract Award (notice to proceed)
ii. From 75% to 10% upon Preliminary Acceptance Certificate (PAC)
iii. From 20% to 50% at Final Acceptance Certificate (FAC)
iv. From 5% to 20% upon achieving the 30-Day Reliability requirement

30-Day Reliability (p 22)

The TTC is restructuring its milestone payments. Included in this restructure is a larger percentage due upon achievement of the 30-Day Reliably requirement.

Updates:

The TTC has restructured its milestone payments for the next bus procurements. The percentage due upon achievement of the 30-Day Reliability requirement has been increased to 20% from 5%.

Energy Consumption (p 52)