Transportation Deployment Casebook/2023/Vancouver SkyTrain

About the Vancouver SkyTrain
The Vancouver SkyTrain is a metro system made up of three lines which stretch across metropolitan Vancouver, British Colombia, south of the Burrard Inlet. The system opened in 1985 and serves many key urban centres (including Metrotown, Coquitlam, New Westminster, Surrey, and Richmond), Vancouver’s CBD, and YVR Airport. It is an integral part of the city’s transport network and continues to see expansion projects to serve more passengers.

There are two modes used in the SkyTrain network, the Canada Line uses conventional light metro technology while the Expo and Millennium Lines use Linear Induction motor technology from Bombardier/Alstom. Both modes are characterised by their high frequencies, grade separation, fully automated running, and moving block signalling. Both technologies are unified under ‘SkyTrain’ branding. Figure 1 shows a section of the SkyTrain Guideway. ''Figure 1: SkyTrain Guideway as seen from Production Way-University Station on the Millennium and Expo lines. Original Photograph taken 17/2/2023.''

A Brief History of Public Transport in the Vancouver Area
Prior to the SkyTrain’s introduction the initial route from Vancouver CBD to New Westminster was served by the British Columbia Electric Railway company with a streetcar and interurban line. The streetcar line was closed in 1955 and the interurban in 1958 along with many others in North America.

This left Vancouverites primarily reliant on their bus, ferry, and trolleybus networks and private vehicles. A lack of urban freeways in Vancouver coupled with increasing reliance on road-based transport led to a search for solutions.

In 1968, the Vancouver Transportation Study proposed an extensive system of surface freeways to alleviate the city’s transport problems that was adopted by council for a time. This would include a large interchange near the CBD and a new crossing of the Burrard Inlet. The only part of this plan to have been constructed was the Dunsmuir and Georgia Viaducts.

On the other hand, Alderman Harry Rankin composed a pamphlet called “The Case for Rapid Transit” in 1970 arguing against the 1968 plan and advocating for a rapid transit network across Vancouver instead. Rankin argued that freeways would cost more and achieve fewer benefits to the city while cutting up historic neighbourhoods like Vancouver Chinatown.

Initially, a ‘Light Rapid Transit’ (streetcar) reintroduction was considered, however this was rejected in favour of the higher capacity but experimental ‘Advanced Light Rail Transit (ALRT)’ (Linear Induction Light Metro) technology employed on the Expo and Millennium lines. A major attraction of this technology is grade separation meaning it would be independent of other modes, unlike Vancouver’s busses and trolleybuses.

SkyTrain Technology
The SkyTrain is like other automated light metro systems built around the same time, sharing Linear Induction technology with Toronto Transit Commission’s Line 3 (Scarborough RT), and sharing its high frequency with the Docklands Light Railway in London. The system is an evolution of metro technology however it claims to have been the first driverless rapid transit system in the world.

ALRT itself was developed by a Canadian Crown Corporation and thus its selection for the project was politically popular. The technology has not been applied broadly elsewhere however some stand-alone systems have used Linear Induction technology including the JFK AirTrain in New York. While ALRT did not catch on more broadly and was not even used on the SkyTrain Canada Line, other technological aspects of the system have become increasingly common including moving block signalling, driverless operation, and level boarding. ALRT has been used on extensions of the Millennium and Expo lines and will be used on the upcoming Broadway Subway.

SkyTrain Usage
The SkyTrain did not only capture riders of superseded bus services but also appealed to car drivers. Vancouver has a particularly high percentage of trips taken by transit for a North American city at 18.3% (and over 50% of trips were using Transit or active transport modes). The SkyTrain contributes significantly to this, enhancing other modes such as busses and ferries, and on its own it has one of the highest utilisation rates of passengers per kilometre per year at ~2,029,358pax/km/year in 2019.

In the financial year 1987/88, Skytrain recorded a ridership of 998,832pax/km/year, in 1995/96 after the opening of the Surrey Extension this jumped to 1,354,960pax/km/year, over a 35% increase in pax/km/year showing major growth in utilisation as the system both expanded and became more established and normalised as a mode of travel.

Ridership has been assisted by policy, for example, the lack of freeways discourages car use on trips to the city or between many urban centres. Additionally, complimentary high-frequency bus services and feeder busses (Rapid Bus and B-Line) have expanded the catchment areas of stations particularly in Surrey providing the SkyTrain with a new market of customers. The 99 B-Line was so successful as to be partly replaced by a SkyTrain extension.

SkyTrain Growth, Development, and Adaptations - 1985 to Present
The growth and success of the Vancouver SkyTrain since 1985 has come from a combination of public policy and private sector development.

Continual investment in improvements to the system including new rolling stock, station upgrades, and new and extended lines from government has kept the mode competitive and built ridership. Rolling stock has also improved over time, air conditioning, open gangways, and automatic doors have been employed increasing passenger comfort and vehicle capacity while speeding up boarding. Additionally, the introduction of the compass card in 2013 (smartcard technology) made riding the SkyTrain more convenient.

Neighbourhoods near SkyTrain Stations have increased in density since the opening of the Expo, Millennium, and Canada Lines. Private sector investment especially in housing and commercial space near stations such as Metrotown and Main Street/Science World which has significantly driven up ridership. Developments integrate with stations, for example the Over Station Development (OSD) at New Westminster provides many restaurants and other businesses within 100m of the station entry with many being directly level with the platforms like at Chatswood in Sydney.

New lines and extensions such as the Evergreen Extension of the Millennium Line and the Canada Line have seen integration between the public and private sectors. Public sector investment in the line resulted in high quality stations and services such as 4G connectivity in the Burquitlam Tunnel. The private sector has contributed by developing new housing and new businesses at stations. At Inlet Centre where previously there were much lower intensity land uses high density housing has been constructed and at Moody Centre Industrial warehouses now host several craft breweries. There is an important time factor with this, the Expo Line has higher densities than the Millennium as the latter was constructed a decade and a half after the former.

Vancouver has flirted with other modes post-SkyTrain introduction. Initially the Evergreen Extension was to be a Light Rail (tram) line to not compete with West Coast Express patronage. During the Winter Olympics in 2010, the ‘Olympic Line’ (a demonstration tram line) operated on a route from Olympic Village Station to Granville Island and was tipped to be extended to Waterfront Station. This line failed to secure funding from Translink as a lower priority project compared to SkyTrain extensions such as the Broadway subway.

Quantitative Analysis of SkyTrain Ridership Trends
Historic data provided by Translink is limited to their ‘conventional system’ trips and boardings for each year back to 2002. This includes SkyTrain but also Bus, Trolleybus, SeaBus, and West Coast Express services. To estimate SkyTrain patronage as a percentage of this number, the average percentage of monthly trips taken by Skytrain was calculated based on the 2022 data available, then it was applied back to 2002 assuming the same rate to get an approximate value (SkyTrain was 36% of the 'conventional system' in January to November 2022). This was checked for plausibility vs the SkyTrain data provided to JRTR in 1998 and seemed plausible. This, however, will reduce the accuracy of the results generated.

Data was analysed using Ordinary Least Squares Regression in three sets, one factoring in the COVID-19 Pandemic years 2020-21, one which did not factor data from these years into the analysis, and one that did factor the pandemic but assumed that there will be no recovery beyond the 2019 ridership. Data was fit to an S-curve and the inflection point was noted. For COVID-19 Data, SMax (Saturation status level) was calculated to be 297,500,000 Pax/Year and ti (inflection year) was 2020. For Non-COVID-19 Data, SMax was calculated to be 190,000,000 Pax/Year and ti was 2005. For the SMax at 2019 set, SMax was assumed to be 161,536,941 Pax/Year and ti was 2002.

The S-curve was calculated by the Equation 1, considering SMax, ti (inflection year), t (the current year), and b (estimate coefficient). The best curve was selected for each set using the highest R-Squared (coefficient of determination) ≤ 1.

$$S(t)={\frac{S_{Max}}{[1+e^{-b(t-t_i)}]}}$$ Equation 1

COVID-19 Data
Considering the COVID data (Figure 2), the S-curve is a poor visual fit to the rest of the recorded data due to the two strong outliers in the 2020 and 2021 results. By this metric, the mode has just reached the inflection point and is in the growth stage. This is a poor fit for the data and as a result the outlier years should not be considered when examining the lifecycle of the SkyTrain as the lower ridership results were not due to any change or issue with the technology itself but the broader global impacts of COVID-19.

In this data major system events noted in the COVID-excluded data can be seen but significantly the impact of COVID is obvious in the 2020-21 data.

''Figure 2: COVID-19 years included. Data from Translink and JRTR. The S-Curve in this data has an R-Squared of 0.574951201.''

Non-COVID-19 Data
Considering the non-COVID data (Figure 3), we can identify that the mode is in its growth phase but nearing the mature stage. The birthing phase would be from 1987 (launch year) to around 2000 just before the Millennium Line opened in January 2002. There is still potential for the growth stage to continue with projects such as the Broadway Subway, however the overall system expansion rate after that has slowed meaning that entry into the mature stage is likely.

In the measured data for Figures 2 and 3 the following events can be identified: The opening of the millennium line (impact between 1996 and 2002), the Canada Line in the 2009 to 2010 jump, and the Evergreen Extension (in 2016). These openings contribute to a broad upward ridership trend overall, however after the initial opening hump ridership settles to a slower rate of increase.

''Figure 3: COVID-19 years excluded. Data from Translink and JRTR. The S-Curve in this data has an R-Squared of 0.956915239.''

COVID-19 Included, Assume SMax Was Achieved in 2019
While ignoring COVID-19 years when considering the growth stage of the technology provides a more accurate picture of SkyTrain’s lifecycle, the exclusion of this data means a picture of the post-COVID recovery is not possible with this analysis.

To achieve a better line of best fit, figure 4 plots the S-curve assuming SMax was equal to the maximum measured ridership in 2019. The result produces a better inflection point at 2002 but would place the technology in the mature phase of the lifecycle with the potential for continuing poor ridership to class it as declining. Such a result could be possible depending on COVID-19 recovery trends. From the available 2022 data, SkyTrain trips have recovered to 66% of 2019 levels (January to November), as compared to 49% in 2021, however the impacts of COVID-19 are still ongoing.

''Figure 4: COVID-19 years included, SMax taken as 2019. Data from Translink and JRTR . The S-Curve in this data has an R-Squared of 0.184149387, lower than Figure 2, however it has a better visual fit.''

Conclusion
Overall, the model using non-COVID data and an above 2019 SMax (Figure 3) is the most appropriate for assessing the lifecycle stage of the Vancouver SkyTrain. This is because future system expansions and developments are in planning and considering 2022 SkyTrain trip data an eventual recovery to pre-COVID conditions within the decade seems plausible.