Phoenix to Tempe Regional Rail

For a street-running light rail line, Valley Metro light rail outperforms. But its greatest flaw is its speed. Transit becomes the default option when it becomes faster than driving, and in Phoenix that requires grade-separated right of way.

AZDOT’s planned commuter rail would parallel the light rail’s core route between downtown Phoenix, Sky Harbor, and downtown Tempe. However, it will be a low frequency, lower ridership service requiring massive capital for any reasonable initial operating segment, and it won’t connect well to existing density and transit lines.

One rail line can solve both problems and more. Light rail vehicles would operate on an initial operating segment of a future regional rail network, and the project might be cheaper than currently planned. It’s the fastest and least expensive way to halve the transit trip time on an already frequent and high-ridership transit service between 3 of the region’s largest, densest, most parking-limited job centers, making transit the fastest way between downtowns.

Transit plans between downtowns have 3 problems

1. Light rail between downtown Tempe and Phoenix is too slow, especially because much of the demand is from one downtown to the other or to Sky Harbor, not to other intermediate stops.
2. Planned commuter rail will require massive capital expenditure and result in low frequency, comparatively low ridership service with poor connections to light rail and job centers that won’t improve current riders’ commutes.
3. Because the light rail tries to move riders through the corridor quickly, it skips multiple destinations and bus connections. If it stopped more frequently, it may spur more transit oriented development along the corridor.

light rail Sky Train streetcar planned commuter rail freight right of way

3 moves solve these problems

1 Problem: Light rail is too slow
Solution: An initial commuter rail operating segment on new track along the existing freight right of way could span the downtowns with short connections to existing light rail at either end, so light rail vehicles could immediately activate the corridor with frequent express service between downtowns, speeding current light rail riders’ journeys.¹

2Problem: Commuter rail isn’t well connected
Solution: Once expanded past the initial operating segment, commuter rail can run on the same tracks as light rail between downtowns. Shared stations mean seamless connections between commuter rail and light rail, and a new station at Central Ave best serves the stadiums and downtown Phoenix and could connect to the N-S light rail line.

3Problem: Existing line doesn’t serve local trips well
Solution: A new service will operate on Washington and Jefferson streets. Four new infill stations can better connect to bus lines and local destinations, now that people traveling between downtowns have a faster option. The line could extend for increased frequency in Tempe or as a branch of the Tempe Streetcar. An extension is one of many future improvements made possible by the rail line, discussed below.

Rail becomes the fastest way to get between Phoenix, Sky Harbor, and Tempe

Today, light rail speeds are limited by the speed limits of the streets they run on and limited transit signal priority still leaves many trains waiting behind red lights. With a grade-separated right-of-way and minimal intersections—all with traffic preemption and none with traffic volume concerns—light rail could consistently reach 55mph. Forcing transfers to commuter rail wouldn’t have the same benefit.¹

With light rail this fast between downtowns, transit is the fastest option, period.³ People who live in one downtown and work in the other, ASU students who would otherwise take the ASU shuttle between campuses, Suns fans using a park & ride to avoid traffic, and business travelers getting between the airport and the convention center no longer need to weigh speed versus cost; transit is always the best value.

The utility of fast transit will convince some people living in downtown Phoenix to go car-free, some families to keep 1 car instead of 2, and many more to take transit to the airport for vacation. In turn, the trunk will directly and indirectly lead to significant ridership increases and make transit a more important part of the Valley of the Sun’s transportation infrastructure.

Today, light rail operates like two systems, with few riders between downtowns

Routing light rail on the commuter rail tracks speeds and integrates the network

Running light rail on an initial commuter rail segment means better service and more ridership metro-wide, even before the first commuter train is ordered

Faster light rail service will benefit all rail transit trips and trips between the east and west valleys with faster, more reliable service:

• With faster transit between the east and west valley, residents gain access to more jobs, becoming new daily riders
• Future commuter rail riders gain cross-platform transfers to light rail for better access to job centers
• Bus lines feeding the rail, especially near the new faster track, will see increased ridership and transfers
• Riders on all parts of Valley Metro rail see improved service reliability because there are no traffic lights to get stuck behind on the trunk
• Faster service is less expensive to operate, because faster trip times mean an operator can cover more distance in the same time, meaning resources can be deployed elsewhere
• More office, residential, and retail will locate near rail stations, and their inhabitants will be more likely to use rail, contributing to ridership increases

When regional rail is expanded past the initial operating segment, benefits compound.

New investments should double down on the region’s most productive service: light rail⁵

Integrating light rail makes commuter rail less expensive

Not much needs to change to allow light rail on the planned commuter rail initial operating segment.⁸ On net, this plan likely saves money and is easier to negotiate with Union Pacific versus the existing commuter rail plans. A preliminary plan is here.

Additional expenses:

• Full double track between Phoenix and Tempe, which allows frequent, reliable service.
• Two new stations with 4 tracks, allowing separate platforms for light rail and commuter rail.
• ~1,700 ft of additional track to connect existing light rail to the initial commuter rail operating segment.
• Marginal cost of provisions for- or installation of catenary which feeds both commuter and light rail, and/or light rail vehicle power and signaling alterations

Savings and simplified negotiations:

• There’s no need for another rail bridge over Tempe Town Lake, because commuter rail can use the existing light rail bridge. This savings likely more than makes up for all the above expenses. If a commuter rail alignment along the 202 and 101 around downtown Tempe is selected, this plan still provides improved service to downtown Tempe.
• Without the need for a new bridge, the commuter rail also doesn’t need to cross the existing UPRR tracks north of Tempe Town Lake. Eliminating that crossing greatly simplifies negotiations with the railroad and the FRA (or removes a costly grade separation); simplifies operations and improves reliability; allows for station placement closer to downtown Tempe, Mesa, and Gilbert; and likely reduces the number of freight rail crossings farther southeast. Compliance with UPRR’s preferred 50 ft between passenger and freight tracks would be unchanged, and grade separations and crossing closures can still be used as negotiating leverage.

This new rail line makes future improvements possible

Here’s how we can start

1Safeguard rights-of-way⁷
~6.7ac will need to be acquired from private owners. Most urgently, the parking lots between Chase Field and 1st Ave, much of which could be redeveloped after project completion. Parcels of interest are shown on this map.

2Update commuter rail plans with this initial operating segment
This line can improve light rail service immediately and jumpstart a larger commuter rail network over time. Ensure compatibility with light rail and commuter rail, and potentially plan for future light rail rolling stock to have higher top speeds.

3Engage the FRA to begin a Petition for Approval of Shared Use
The FRA seems poised to grant approval of shared use in a case like this.

Footnotes

---

1. Can’t people just transfer to commuter rail?

No, without integrated light rail, none of the problems outlined at the top are solved. Commuter rail is planned every hour (every half hour at peak), so riders on up to 4 out of 5 light rail trains per hour (assuming 12 min headways) have no commuter rail train to transfer to. But commuter rail frequency could conceivably increase.

More fundamentally, the problem is transfers. It takes 3 minutes optimistically to walk between the 3rd/Mill light rail station and Tempe commuter rail station, and another 5-10 minutes optimistically to walk between stations in DTPHX, depending on station location. Assuming at least 3 minutes in the schedule at each transfer to account for slower walkers and slightly delayed trains, that’s at least 14 minutes of walking and waiting, fully cancelling out the speed benefit. And that assumes trains aren’t more than a few minutes delayed and that travel times allow for timed transfers at multiple points on the same lines. Even if a rider’s start or end point is in one of the downtowns, they’re still likely to need to walk because both commuter rail stations are on the periphery of their downtowns.

If and when this project improves service to the point that ridership doubles or triples and trains are much more frequent, I’ll be the first to advocate for spending more to create seamless transfers between deinterlined services. With that kind of ridership, businesses start responding to transit station locations when they choose offices, rather than the opposite currently.

---

2. Travel time estimate: average alternative 2 travel time between Mill Ave/3rd St and Chase Field/3rd St estimated to be 11m 49s, 1.95x faster than existing light rail. Additional schedule padding may need to be added to this estimate. Full estimated travel time matrix below:

Rotate phone to read table

		Existing EB	Existing WB		Alt 2 EB	Alt 2 WB	Avg % improved		Alt 1 EB	Alt 1 WB	Avg % improved
Mill Ave to 44th		9m 34s	10m 34s		5m 13s	5m 11s	194%		5m 30s	5m 24s	185%
44th St station		26s	26s		26s	26s			26s	26s
44th to 3rd		12m 30s	12m 30s		6m 10s	6m 12s	202%		6m 17s	6m 20s	198%
3rd St station		26s	26s		26s	26s			26s	26s
3rd to DT Hub		2m 34s	2m 34s		1m 46s	2m 0s	136%		1m 46s	2m 0s	136%
Total: Mill Ave to Chase Field		22m 30s	26m 30s		11m 49s	11m 48s	195%		12m 13s	12m 10s	189%

Assumptions: Speeds based on curves and other speed limit attributes of designed alignments, listed on click here. Assumes 4″ superelevation where possible and 2″ unbalanced superelevation, based on desirable upper limits from Valley Metro LRT Design Criteria Manual (p18). Assumes 3mphps acceleration and deceleration and 55mph top speed based on maximum operational characteristics of Valley Metro’s old and new light rail vehicles. Increased top speed of future rolling stock could reduce travel times further in the future. Assumes station dwells of 26 sec based on average station dwell data collected Fall 2024. Existing light rail travel times are a synthesized average of Google Maps and Valley Metro schedule data accounting for each each interstation and route timing as a whole.

A similar diagram in ADOT’s Passenger Rail Corridor Study (p77) mirrors the findings above. It calculates 8 min travel time between Phoenix Union Station and Sky Harbor, and 6 min travel time between Sky Harbor and Tempe assuming diesel-electric engines and including station dwells.

---

3. Travel time comparison: 17 min drive versus <12 min proposed light rail ride between Mill Ave/3rd St and Chase Field/3rd St. More example trip approximate comparisons below:

Example trip	Stations	Existing rail	Proposed rail	Drive time assuming no traffic	Fastest mode?
Downtown hopping, now drawn closer together	Mill Ave/3rd St to 3rd St PHX	23m	12m	17m	Rail
ASU student traveling to DTPHX campus	University/Rural to Van Buren/Central	34m	22m	30m via ASU shuttle*	Rail (vs ASU shuttle)
Suns fan using a park & ride to get to a game	Price 101/Apache to 3rd St PHX	42m	28m	19m	Depends on gameday traffic
Convention visitor arriving from the airport	Term. 4 Sky Train to 3rd St PHX	26m†	15m‡	12m	Depends on traffic, rail freq., ridehail availability

* Other considerations: ASU Maroon shuttle boards on the other side of Tempe campus. Light rail is more frequent, has longer hours, and is likely more reliable. Both are free to ASU students. The same route would be a 17 min drive in a car. If ASU can rely on light rail for transport between campuses, there may be a chance of getting funding for improvements or operations from them.

^† Existing rail assumes 5 min Sky Train ride, 5 min transfer, 3 min wait (based on Sky Train frequency of 5 min), and 13 min light rail ride

^‡ Proposed rail assumes alternative 2c, including a new Sky Train station: 4 min Sky Train ride to new station, 2 min transfer, 3 min wait (based on Sky Train frequency of 5 min), and 6 min light rail ride

Existing light rail travel time calculated via Google Maps. Proposed light rail travel time synthesized from Google Maps and from alternative 2 in the above method. Drive time calculated via Google Maps, assuming no traffic.

---

4. Access estimate: The trunk line increases jobs and people accessible within the same transit trip time. The following estimates are underestimates, because ASU is not counted as a “workplace” for students, and ASU students without a job are not included in the population count, among other approximations listed below:

	Veteran’s Way/College	Van Buren/1st Ave	Sky Harbor/44th Street†	Chase Field/3rd St†
Jobs within 30 min on transit: existing	259k	274k	363k†	254k†
Jobs within 30 min on transit: alt 2b	364k	380k	461k	391k
% increase accessible jobs	141%	139%	127%†	154%†
People* within 30 min on transit: existing	78k	68k	79k†	40k†
People* within 30 min on transit: alt 2b	89k	85k	98k	77k
% increase accessible people*	115%	125%	125%†	193%†

To estimate access: isochrones from Travel Time—assuming existing Nov 2024 transit service, then combining isochrones for each proposed station (assumes alt 2b) with estimated interstation timing—were mapped onto 2022 US Census job and population* estimates from OnTheMap.

Estimates are approximate:

• ASU is not counted as a “workplace” for students, and ASU students without a job are not included in the population count
• Existing access estimates overestimate access from north of the DTPHX Hub, because Nov 2024 service doesn’t require a transfer between N-S and E-W lines, whereas soon that transfer will be necessary until the trunk is constructed
• Proposed access estimates slightly overestimate access from east of Mill Ave, because they don’t account for an added transfer to the streetcar for access to stops between Mill and Sky Harbor
• Combining isochrones with estimated interstation timing does not exactly reflect transfer times between rail and other modes.

* “People” refers to the home address of people working in the state. Notably, this excludes ASU students who don’t work in addition to being a student.

^† Existing measurements come from existing stations (44th St/Washington and 3rd St/Jefferson), whereas proposed come from proposed location of trunk stations. This undercounts access with the trunk, because trunk stations are in addition to, rather than replacing, existing stations.

---

5. Ridership comparison: Commuter rail ridership figures are projections for 2030 from MAG’s Commuter Rail System Study, 2010. P3-24 lists full system daily weekday ridership at full buildout. The “Grand to SE and Yuma to Tempe” alternative (most likely scheme; 12,200 daily riders) was adjusted by 48%: post-COVID ridership recovery of peer systems based on 2024 versus 2019 total ridership. Peer systems were any systems mentioned in the study: Sounder in Seattle (46% ridership recovery), North Star in Minneapolis (17%), Front Runner in Salt Lake City (79%), WES in Portland (31%), Coaster in San Diego (64%), Metrolink in Los Angeles (52%), Caltrain (39%) and ACE (51%) in the Bay Area. The average of recovery figures is 47%, whereas the recovery of total ridership of all systems summed is 48%.

All other ridership figures are average daily weekday ridership from Valley Metro’s 2024 annual ridership report (hidden sheet: “Graph Data Sheet”). Potential BRT corridors are those identified in MAG’s Regional Bus Rapid Transit Feasibility Study, 2021.

---

6. Cost estimate and comparison: Alternative 2 is estimated to cost $499m, not including contingency, ~6.7ac of land acquisition, or any compensation necessary to Union Pacific outside replacing infrastructure:

	LRT trunk alt 2a	LRT trunk alt 1	Streetcar across Mill Ave Bridge
Total cost (inflation adjusted)	$249m	$262m	$29m
Cost per mi (inflation adjusted)	$32m/mi	$34m/mi	$37m/mi
Total cost (inflation + transit costs increase adjusted)	$499m	$523m	$59m
Cost per mi (inflation + transit costs increase adjusted)	$64m/mi	$67m/mi	$74m/mi

See the allocations and adjustments broken out by standard cost category here. Here’s how cost estimates were derived:

1. Cost categories: cost per unit calculated for each of the FTA’s standard cost categories based on Valley Metro’s initial operating line using the FTA’s Capital Cost Database
2. Manual adjustments: itemized costs were adjusted where noted here to account for different contexts
3. Inflation adjustment: costs were scaled to account for inflation, from 2008 to 2024, using the BLS’s CPI Inflation Calculator
4. Transit costs increase: costs were scaled again by 200% to account for higher construction costs on recent transit construction projects in the US generally. 200% was chosen as approximate based on more recent Valley Metro project costs per mile (Central Mesa Extension, South Central Extension, Northwest Extension Phase II), adjusted down from an actual increase ~250% due to extension projects being more complex or including more bridges per mile than the initial route. More details here. Costs are shown before and after this multiplier.

In addition to the above, estimates do not include:

• Moving sidewalk connection to Sky Train (Alt 2b) or new Sky Train station (Alt 2c)
• Potential future stations: Trunk at Center Pkwy, Trunk at Phoenix Rising Stadium, or Washington/Jefferson streetcar infill stations
• Related projects: Ped/bike/service vehicle over/underpasses at rail/canal crossings, 150 ft trench through Sky Harbor, 24th St grade separation, etc.
• Added costs for more stringent commuter rail standards regarding electrification, signaling, track structure, and larger/more amenitized stations, except when comparing costs to commuter rail costs, as then these differences are built in.

Alternative 2a is estimated at around 25% the cost/mi of the two most recent Valley Metro extensions. Comparisons are approximate: trunk line estimate excludes right of way and contingency.

	Cost (year built)	Cost (2024)	Cost/mi (2024)	Cost/mi compared to initial segment	Notes	Alt 2a cost/mi comparison
Initial Light Rail Segment	$1,315m (2008)	$1,888m	$96m	100%		67%
Central Mesa Extension	$197m (2013)	$265m	$86m	89%	Expect much less expensive per mi: likely wider roads (less expansion necessary) with fewer utility issues; no bridges; fewer intersections/stations per mi	75%
Northwest Extension Phase II	$401m (2023)	$413m	$258m	269%	Expect much more expensive per mi: far more aerial/mi and an elevated station; likely more stops/mi	25%
South Central Extension	$1,345m (2023)	$1,385m	$252m	263%	Expect a bit more expensive per mi: no new bridge across the Salt River necessary but work required on RR underpasses and more work downtown/with track in operation per mi	26%
Trunk Line Alt 2a Estimate	~$499m (2024)	~$499m	~$64m	200% then adjusted		100%
Est. high end Scottsdale/Rural BRT	$600m (2021)(p59)	$694m	$45m		Not used to estimate transit costs increase, only for comparison between potential projects competing for funding	143%

---

7. Right of way acquisition and negotiation: In addition to negotiating with Union Pacific for use of its corridor, this proposal requires ~6.7ac of private land and some coordination between government and quasi-governmental organizations. Requirements are listed below in order of importance (parcels of interest are shown here):

1. Use of Union Pacific’s existing rail corridor
2. All of 7 parcels, all currently surface parking, totaling 3.08ac, well over half of which acreage can likely later be redeveloped
3. Portions totaling ~.29ac, all currently surface parking, from 2 parcels totaling .86ac
4. Slivers of 7 parcels totaling 3.29ac (alt 2) or 3.33ac (alt 1), which should not require removal of any buildings
5. Shared use of the fire lane south of Chase Field, currently owned by Maricopa County
6. Vacant slivers (alt 2) or portions of surface parking (alt 1) on 2 parcels of airport land, currently owned by the City of Phoenix.
7. Vacant slivers of land on 5 parcels currently owned by AZDOT, SRP, and the City of Tempe. A small, vacant portion of a 6th parcel owned by SRP would be used for the streetcar extension across Mill Ave Bridge

---

8. Shared light rail and commuter rail track:

a. FRA regulatory compliance: Ordinarily, FRA rules don’t allow light rail transit vehicles to share track with commuter and freight rail. I’ll explain solutions to both separately below, acknowledging my limited knowledge of FRA regulations.

i. Freight: Though most of the initial operating segment would be located within the Union Pacific right of way, there would be almost no track sharing. The only potential overlap is at up to four crossovers needed to access to industrial sidings, only one of which looks used today.

Remedy: First, analyze whether the benefits and costs of maintaining these industrial sidings outweigh the regulatory and operations burdens necessary to keep them open. If they do, freight could be “temporally separated” by operating only in the early morning before light rail begins operating. This is common practice on UTA’s TRAX light rail, among other US systems. Other agencies’ experiences, additional FRA guidance (see table at the bottom), and the limited instances of shared track should make this process straightforward.

This proposal eliminates a crossing of the through freight tracks present in current commuter rail plans. That crossing likely wouldn’t allow for temporal separation. Negotiations with the FRA could be vastly simpler without it, even when accounting for negotiations regarding the next section…

ii. Commuter rail: This proposal assumes light rail and commuter rail trains would share tracks at the same times of day (“simultaneous joint use”) on the initial operating segment. This joint use would happen only after the commuter rail system was expanded past its initial operating segment and commuter trains started running on that longer line. Therefore, any additional time necessary to get approval of joint use should not impact the opening of the initial operating segment by light rail vehicles.

Remedy: The FRA gave guidance in 2000 on what would be necessary to allow simultaneous joint use. Regarding European systems which operate with simultaneous joint use, like Karlsruhe’s Stadtbahn, the FRA said, “If some of those systems were replicated in the United States in every detail, FRA would very likely approve them by rule or waiver.” Here are the differences they cited at the time, and how they’d be addressed in this case ranked roughly by impact:

• Karlsruhe’s trains operate with Automatic Train Control. In the US, a rough equivalent, Positive Train Control, was mandated in 2008 and would be installed on any commuter rail system in Phoenix.
• In Karlsruhe the predominant traffic under simultaneous joint use with light rail is scheduled passenger trains, rather than freight. In Phoenix, all traffic under simultaneous joint use would be scheduled passenger trains. Any freight would have temporal separation, as discussed above.
• Karlsruhe’s light rail and mainline rail equipment differs less in mass and structural strength than the same in the US. Recent FRA Alternative Crashworthiness Compliance standards have substantively standardized the mass and strength requirements of mainline rail cars between the US and Europe. Newly allowed commuter rail trains, likely candidates for Phoenix commuter rail vehicles, are often approximately the same weight per car length as Valley Metro’s existing light rail vehicles. All would be much lighter than freight cars or locomotives:
  • 1.12k lbs/ft: Valley Metro’s S700 light rail vehicles at 102.5k lbs for one 91.5 ft car
  • ~1.16k lbs/ft: Stadler’s FLIRT single level, low floor EMU at ~240k lbs for a ~207 ft train
  • 1.15k lbs/ft: Stadler’s KISS bi-level EMU at 592k lbs for a ~515 ft train
• Karlsruhe has limited grade crossings, each with a four-quadrant gate. In Phoenix, there are 13 at-grade crossings on the corridor, 3-6 of which would likely be closed or grade separated during construction. It may be possible to fit the remainder with four-quadrant gates.
• Karlsruhe’s light rail vehicles have very high braking capacities, as compared to US light rail vehicles. Valley Metro’s new S700 light rail vehicles achieve service deceleration of 3.0 mphps (1.34 m/s²) and emergency braking of 5.0 mphps (2.24 m/s²). That compares to 1.6 m/s² for Karlsruhe’s GT8-100C/2S and GT8-100D/2S-M, both operating in 2000.
• All trains in Karlsruhe’s network use a common communications system, operate under the same operating rules, have integrated train crews trained to operate all types of vehicles and limited to 40 hour work weeks, and have centralized dispatching. Similar standards should be possible in Phoenix if set as a goal when commuter rail operating procedures are created.

The FRA ends their guidance reiterating the need for PTC and vehicle weight changes, then without even knowing that those barriers would be removed 8 and 18 years later, say, “However, we are open to consideration of any reasonable proposal.” This proposal is the perfect case for the FRA to approve a “Petition for Approval of Shared Use” for light rail and commuter rail simultaneous joint use.

b. UPRR negotiations: No part of negotiations with UPRR should be made harder by combining light rail and commuter rail, as compared to the current commuter rail proposal (schematic starting p 55). In fact, reusing the light rail bridge over Tempe Town Lake would remove all major UPRR/transit crossings from the initial operating segment, removing potentially the most challenging issue to negotiate.

The existing plan contemplates a new commuter rail bridge over Tempe Town Lake west of both existing rail bridges, likely because constructing the new bridge in the 50 ft space between the two would be exceedingly complicated. To access that bridge, commuter rail would need to cross existing freight tracks to the north of the lake over a costly grade-separated crossing or at-grade. An at-grade freight rail/commuter rail crossing would be hard to negotiate with UPRR, would likely result in multiple-minute-long commuter rail delays (because UPRR would not concede on scheduling), and would make FRA negotiations harder (because the two likely could not be temporally separated). This plan would eliminate the need for a freight rail/commuter rail crossing because commuter rail could use the existing light rail bridge, solving those problems; allowing for station placement to the north and east of freight tracks, closer to downtown Tempe, Mesa, and Gilbert; and reducing the number of freight rail crossings between Tempe and Eloy from ~6 in-use sidings and 2 branches to ~5 in-use sidings and 1 branch.

The corridor generally has enough room for transit and freight tracks to be separated by 50 ft (UPRR’s preferred distance, p L2)—no change with light rail added. Even if UPRR installed a continuous second freight track, there could generally be 33ft+ of separation. That’s more than LA Metro’s 20ft negotiated with UPRR recently for the West Santa Ana Branch (via X), or between Valley Metro and the same UPRR line in downtown Tempe, ~25ft with no crash fence (both light rail). More examples of and info on shared transit and freight rail corridors. Agencies also have leverage: projects like the 24th St overpass or Grand Canal grade separations, potential rail crossing closures, improved at-grade crossings, new freight tracks.

c. Operations: Interlining services necessarily increases complexity. In the case of this proposal, that complexity can be limited.

i. Timetabling with vehicles going different speeds: Valley Metro’s S700 light rail vehicles accelerate faster (3mphps) than FLIRT and KISS trains (1.7-2.7mphps) (two representative high-performance EMU commuter rail vehicles) but have a slower operational top speed (55mph vs 99-124mph). This means faster commuter trains could get stuck behind slower light rail vehicles on the corridor, getting them off schedule. With separate tracks and platforms for the two services at Sky Harbor station (discussed below), commuter rail trains could pass light rail trains there. This means the largest travel time differential would be 1.5 min between DTPHX and Sky Harbor. In a vacuum, a 1.5 minute difference is easy to timetable around at frequencies no tighter than every 12 minutes (the highest frequency contemplated for light rail today), but street-running light rail stuck at traffic lights will never be as reliable as commuter rail. On the other hand, 12 minute frequencies mean if a commuter rail train was scheduled 10 minutes after the previous light rail vehicle and 2 minutes before the next one, the first light rail vehicle would need to be exactly 10-11 minutes delayed to slow the commuter train. With some dispatching help, this should not be a problem. Higher frequencies might necessitate riskier timetables, but should also justify more infrastructure and service investments to make light rail more reliable and to limit potential for delay on the corridor. Valley Metro might also consider buying or upgrading light rail vehicles with higher top speeds to reduce the differential and improve travel times further. Importantly, existing commuter rail plans contemplate a freight rail/commuter rail crossing north of Tempe Town Lake, and Union Pacific is unlikely to be willing to help commuter rail trains stay on schedule. This crossing would introduce the risk of multiple-minutes-long delays, potentially worse than the possibility of easier-to-mitigate light rail/commuter rail scheduling conflicts.

ii. Crush loads with trains of different capacities: Sometimes, a smaller vehicle operating on the same route as a higher-ridership line with larger vehicles can see crush loads. In this case, 1) light rail on this corridor is generally no where near crush loads at any point, newer trains have significantly more space than older ones, 3-car trains can improve capacity by another 50%, and increased frequency could 3x capacity or more; 2) the two services would only share 3 approximate station locations, so most people would not choose the next train agnostic of service; 3) smaller commuter trains have a similar capacity to 3-car S700 light rail trains, and 4) if more frequent 3-car light rail vehicles were consistently hosting crush loads, more infrastructure and service investments would be justified.

d. Infrastructure: Commuter trains and light rail are able to operate on the same tracks. They both use standard gauge track and can fit under the same catenary wires. There are three places where accommodations are necessary:

i. Electrification standards: Valley Metro streetcar and light rail operate at 750V DC. Stadler FLIRT commuter trains (as a representative example) are spec’d for alternating current (15kV/16.6 Hz or 25 kV/50 Hz) or much higher direct current voltage than light rail (3 kV DC 0r 1.5 kV DC).

Potential remedies: 1) buy commuter trains that run on 750V DC (as the Long Island RR does, though theirs operate via 3rd rail), 2) buy or retrofit either commuter or light rail trains to run on dual voltages (relatively common, as seen in New York or Karlsruhe), or 3) run diesel- or battery multiple unit commuter rail trains.

ii. Vehicle and platform dimensions: Commuter trains are usually slightly wider and have floors slightly higher than Valley Metro’s light rail trains. Though they have no problem sharing track designed for both, they can’t easily share platforms. The likely solution is to split each track into two and have platforms at separate heights and distances from track centers, as shown below. This configuration would only be necessary at two stations: Sky Harbor and between Central Ave and 3rd St near the stadiums downtown. It would have the additional benefit of allowing timed overtakes, if necessary.

iii. Rolling stock weight: Existing Valley Metro light rail trains are approximately the same weight per linear foot, ~1.12k lbs/ft, as the FLIRT, ~1.16k lbs/ft, and KISS, ~1.15k lbs/ft, both representative example commuter rail trains (all weights are of empty rolling stock unless noted). Loads are significantly more concentrated on the commuter rail rolling stock (light rail 17k lbs/axle, FLIRT 30k lbs/axle, KISS 25k lbs/axle unloaded or max 52.1k lbs/axle at crush loads), which at worst likely would mean upgrading track, not bridge structure.

The rail bridge over Tempe Town Lake was built to hold two 3-car light rail trains with crush load passenger traffic (307.5k lbs unloaded; 440k lbs crush load). Likely, two 3- 0r 4-car FLIRT commuter rail trains (3-car: ~240k lbs unloaded) would fit that standard, or a single train that was twice as long. At most, track work, lighter commuter rail trains (with many additional benefits), and minimal timetable adjustments should allow commuter rail to operate over the existing bridge. If commuter rail capacity becomes an issue in the future, ridership will already justify more investment.

---

9. Washington/Jefferson line possibilities: Once the Washington/Jefferson line becomes a separate line serving local trips, its Tempe end has many possibilities:

1. End service after Center Pkwy station at a new crossover. This option has the highest reliability but doesn’t connect the line with downtown Tempe.
2. Extend service through Tempe just like today’s light rail route. This service would interline with service on the new express connection to DTPHX, meaning frequency on the interlined portion through Tempe would be doubled: a train every 6 minutes. This option maintains connections but introduces potential for delays, though that potential is small at today’s frequencies. This service might short turn at one of the following, ranked by likely value:
   • After Sycamore/Main (layover track 300 ft after), which doubles frequency through all of Tempe and into Mesa
   • After Smith-Martin/Apache (crossover 1700 ft after), which doubles frequency through most of Tempe, including high-ridership McClintock
   • At Mill/3rd or Veterans Way/College (layover track 400 ft after or 100 ft before), which doesn’t provide frequency benefits
   • At Gilbert/Main (existing terminus), which doubles frequency all the way to high-ridership Gilbert/Main station
   • At Dorsey/Apache (crossover 600 ft before), which doesn’t double frequency at high-ridership McClintock
3. Build an extension across one of the Mill Ave bridges (likely possible; empty streetcars weigh the same as a full semi) to connect to the Tempe Streetcar. This option adds connections to more of Tempe but adds cost. It could be constructed when frequencies are too high to interline with light rail as discussed above. Washington/Jefferson and Rio Salado could serve as two branches of the downtown Tempe/ASU trunk, if a single-tracked terminal still works at 2x frequency. If not, Washington/Jefferson could serve as the trunk. The entire corridor would be served by streetcar trains. There should be no problems serving Washington/Jefferson with streetcar trains:
   • Streetcars already operate when not in service on light rail tracks, so there are no track, speed, electrification, regulation, etc. issues
   • Streetcars are within tolerances to load and unload at platforms: trains are 104″ wide (vs 104.3″ for light rail) and floors are at 13.75″ (vs 14″ light rail platforms [fig 2-7] and 14.4″ light rail trains)
   • Streetcars have a capacity of 120 people (vs up to 520 for a 2-car light rail train), but the vast majority of riders will use the new faster service between downtowns and Sky Harbor, leaving quite manageable ridership on Washington/Jefferson. If ridership is too high, Washington/Jefferson could serve as the trunk to double frequency, and therefore capacity