Advice for the Advisory Committee

Today, the Metrolinx Board met in its first public session after being reconstituted.  “No Politicians Added!” should be their advertising slogan.  It was an extremely boring meeting where the staff presented reports we had all read beforehand, only one director asked any questions, and the public session was all over in an hour.  I hope that the Board was more lively in the long private session scheduled to follow.

One item of business was the creation of an Advisory Committee for the Electrification Study, or more correctly, a committee to advise on the terms of reference for the study.

Metrolinx is very proud of the crew they have assembled for this committee, and I can only hope that this group will actually get to have meaningful input.  Metrolinx isn’t big on meaningful input, but you’ve heard all that from me before.

Although I was nominated by the Union Station Revitalization Public Advisory Committee to sit as a representative on this body, and one issue in the draft terms of reference is Union Station, I was passed over.  I will live. 

Metrolinx does not appear to even know about the Union Station project because in an extensive report on GO activities, it wasn’t mentioned once.  It is only the single most important change in the station coming down the pipe because, without it, GO hasn’t a prayer of handling more riders.  Service buildups planned for the GO network cannot be accommodated without the greatly increased capacity and improved station layout.  However, more frequent service likely on electrified lines will strain even the improved the station’s capacity.

As a public spirited citizen, I offer a bit of advice for the advisors for their work and their eventual recommendations to the Metrolinx Board.

  • Learn all you can about GO service plans and especially the projections for future riding.  This will have a huge effect on infrastructure requirements.  Read the past studies available online (thanks to CEO Rob Prichard’s intervention to short-circuit a staff demand for a Freedom of Information request).
  • Read especially the last one examining the infrastucture requirements to achieve the Metrolinx ridership projections for the Lake Shore corridor in 2031.  Read about all the new trackage that is required, not to mention the fact that Union Station cannot hold all of the trains, never mind the passengers.
  • Understand plans for and capacity of the expanded Union Station and the rail corridors east and west of it.  These place an upper bound on the amount of service that can be operated on GO services running to and from the station.
  • Fight to have the study structured so that the really heavy corridors, the obvious candidates for electrification like Georgetown and Lake Shore are done first with their own report.  If there is no case for these lines, there’s not much point in studying a branch that will see three trains a day on alternate Thursdays.
  • Fight to have an interim report with findings from this first phase.  The sooner we have a definitive report on the heavy lines, the sooner we can decide whether to undertake electrification.
  • Make sure that you understand railway technology including the difference between performance for diesel and electric modes, and between locomotive hauled and self-propelled units.
  • Learn where the rail corridors actually go and what real estate, if any, is available to expand them.  It’s hard to run more service (regardless of the propulsion) if you can’t fit the trains on the available space for track.  Don’t study frequent electrified service for a corridor that cannot physically support it.  If a corridor will be expanded, understand the physical effect on neighbouring land use as well as the potential for noise and pollution depending on the chosen technology.
  • Make sure that GO will include a fleet plan for whatever scenarios are studied including the build-up of electric equipment for the converted lines and the redeployment of other engines and cars to routes with planned service expansions.  If electrification is a good idea, don’t put it off claiming that we must get the worth out of existing trains before we convert.
  • Make sure you understand engine emission and noise especially as this relates to various technologies.  Consider the effect of different service designs including local and express trains, as well as lines with additional stops serving local demand.
  • Make up your mind on whether electric power is clean or dirty, and quantify the effect at power plants of diverting generation from onboard equipment (diesels) to offboard (nuclear, wind, hydro, natural gas, hot air).
  • Don’t let Metrolinx constrain the electrification options based on what they already want (or don’t want) to do.  This is supposed to be a study, and studies look at options if only to justify rejecting them.

These are rather aggressive tasks for an “advisory” committee, but that group should strive to ensure that the actual study will be useful and timely.

Meanwhile, I will sit out here writing my blog and sending free advice their way from time to time.  I’m sure my loyal readers will be happy to add their two cents’ worth, but will forego actually billing for their time.

21 thoughts on “Advice for the Advisory Committee

  1. I hope they start using electricity to power these trains. If they commit to doing this one line at a time maybe the Ontario gov’t will OK the construction of the new Nuclear plant so we have enough power for our needs. I think if more of our transit needs are electric that our local area pollution will start to clear up as well. I think most of the different commuter rail lines that use Penn Station are electric in NYC. Electric is the way to go in dense urban areas.

    I was surprised to see that the already studied Union Station revitalization is going to be overseen by Metrolinx. This project needs to be started. I am guessing that the monies for Union Station is different then the monies for the Station Modernization program. I know that the TTC has put the Station Modernization Program on hold because they had to find the funds for their large LRT purchace from Bombardier. Thankfully the Vic. Park Station modernization is already underway and it will have to be funded anyway. I have recently move to Toronto from BC and ever since I arrived they have talked of upgrading Union Station-is it ever going to start? Or are people only making a living studying this much needed project?

    Steve: Actually, Metrolinx’ role in Union Station is only part of the story. GO Transit owns part of the station (the rail corridor and train shed), and is about to acquire more (the west wing’s office space). The City owns everything else including the air rights over the tracks and the space under them.

    The design work for the revitalization of the station and creation of new passenger concourses for GO is quite advanced, a head lessee for the new commercial space has been selected (but not yet named publicly), and all we are waiting for is conclusion of the funding arrangements with Queen’s Park and Ottawa. At this point, an announcement is expected in the fall with construction to start next year.


  2. The notion that we must squeeze the most out of our existing equipment can and should be countered by the reality that the F59/MotivePower+bilevel combo is ticking over nicely, expanding in places like Minneapolis, Dallas-Fort Worth and Vancouver – after all, the GO F59s freed up by the arrivals of the MP40s have already been leaving in twos and threes to new homes.


  3. I’ll add that it is important to understand the upkeep costs of each technology, not just the outlay in capital. For what additional maintenance you may have in looking after overhead catenary, would more be saved from less track maintenance from lighter trains and reduced maintenance requirements per propelled unit? What about savings in fuel consumption? These can be quite substantial, depending on the level of service.

    About power consumption, people like to make doomsday predictions for our power grid by electrifying GO, so I would encourage the advisory committee to take a look at it. I’ve done some calculations myself, and I’d hope Metrolinx, as well as the general public, understands what the demands are, and realize it isn’t as massive as some people make it out to be. The whole network, if electrified, would certainly warrant a new nuclear reactor, but let’s also bear in mind that infrequently-serviced parts of the network may not be justified for electrification, although there may be exceptions.

    For an example of a potential exception; Bowmanville. Despite being expected to only see rush hour rail service, it could very well be more complicated for GO to not have this one station electrified since there are no other diesel trains running on this otherwise all-electric line. If one of the Bowmanville trains break down, and is diesel while all other Lakeshore trains electric, it becomes difficult to deal with getting a replacement. There are nuances in the contexts of operations.


  4. You raise an interesting point Steve. Just what is the capacity of Union station vis a vis all the extensions that GO is studying?

    I can’t see that a major expansion would be possible.

    Steve: The plans for the expanded concourse are intended to address a doubling of capacity. However, Metrolinx, planning in a vacuum as always, projected a larger increase in demand than GO. The two obvious questions are which projection is correct, and what will we do when the station is “full”? I don’t have any simple answers.


  5. I hope that this advisory committee looks at all the aspects of electrification. Certainly the whole network being electrified would be nice, but certain areas may never be electrified. If there are areas that electrification is just not warranted because of cost versus ridership than perhaps a diesel electric locomotive might be employed rather than straight electric.

    The one constant for the committee should be that electrification should be required in built up urban areas. Promotion of less localized pollution and increased fuel savings should prompt Metrolinx to electrify sooner rather than later.


  6. Maybe we should start looking at what to do with Union now, before we do any huge modifications. As much as I think we will be needing North Toronto, I suspect Union will need actual track expansion at some point, and we should figure out how that could happen and allow for it before we spend too much money.

    I suspect any expansion mean going underground (ala the Penn station expansion), and that’s going to mean going down at least two levels from what I know of the structure. Not something you want to have to fit into a newly renovated station built without any regard for it.

    Specifically, I’d like to know what kind of cost we would be looking at to put the electric lines in a new station under the existing trainshed. Six to eight tracks would probably be more than enough for rapid transit frequencies on Lakeshore, Georgetown/Airport/Kitchener and Richmond Hill, and would free up a lot of space in the existing station for VIA and the diesel system.


  7. Going down to a new underground track level at Union is a non-starter. Think about it, there are major restrictions on what kind of grades are allowable in the Union Station corridor. Momentum grades aren’t going to fly because stop-and-go traffic movement is common here. There are existing road under rail grade-separations that are existing on both sides of Union Station, one of which is brand spanking new to the point it isn’t even open to traffic yet (Simcoe). The east is particularly complicated. I really don’t think it can be done. You’d have to blow up an enormous portion of downtown.

    If any new level is going to be added… the only way is up. And even that is horrendously difficult, arguably impossible once this new roof goes up, with the exception of the highline tracks.

    Steve: Another important constraint is in the approach tracks to Union to the east and west. In the name of development, and because planners didn’t think we would ever need all that space for a rail corridor, we allowed the tracks to be hemmed in by condos and that small tower just west of John Street. Now space through these areas is at a premium. Don’t forget that even if it were possible to build another level of tracks, which I doubt, you need to be able to access it, and that means ramps and support structures for which there is no space. It was hard enough just to thread the supports for the proposed Peter Street South pedestrian bridge across the rail yards let alone a full-scale rail viaduct.


  8. Planning to accommodate infinitely expanding train traffic in the Union Station Rail Corridor by adding more tracks is just as foolish as planning to accommodate infinitely expanding car traffic by building new highway lanes.

    The right thing to do is to distribute the traffic between more origin-destination pairs around the GTA, with some east-west GO trains bypassing Union Station altogether and instead going via the CP North Toronto subdivision or the CN York subdivision. This would work better if the urban centres in the 905 weren’t such planning disasters from a transit and walkability point of view, and if CN and CP were more inclined to share their east-west corridors with GO.


  9. Couldn’t the platforms be doubled in length (maybe half expanded to the east, and half expanded to the west), obviously this would cut out through service, but it would also double station capacity. I would assume that the track to the east and west that splits could be compressed some to allow for that to happen (perhaps the trains would then need to go slower due to higher rates of turning?) Already on the west side there are some “extension platforms”, probably to the east not much extension is possible … although not every track would have to be extended to increase capacity, and not every train has to be 12 cars long.

    Steve: There is little room for platform expansion (nearby space is occupied by the switching tracks needed to route trains to and from platforms), but the more critical problems lie in overall station capacity for pedestrian traffic between trains, platforms, concourse and surrounding buildings. There is also a problem that if we make trains longer, platforms at stations along the line also must be expanded (please everyone, don’t waste my time with proposals for offset stops). Yes, some trains could be shorter than 12 cars, but they tend to be on routes that don’t run very often and therefore don’t consume much of the total capacity of the station. The real problems are on the Lakeshore, Milton, Georgetown and Richmond Hill corridors, all of which are slated to have very frequent service with long trains.


  10. The one “downtown approach” we’re not using is from the south – i.e. ferries. Vancouver’s integration of marine transit should be a guide here – especially when looking at service from Niagara Region since the ferries’ direct routing would help them compete in a way Admiral Giambrone’s Scarborough proposal would not.

    For example, Grimsby is 84km by land, Port Dalhousie 111km, Niagara on the Lake 130km – but each is about 50km from Toronto by water (try it in Google Earth) depending on how they would round the Islands and where they would dock – while many may object while citing Rochester, that was a massively different application – a look at a map (with a border line) would show why.

    From an integration point of view I wouldn’t preclude GO trains from Niagara – Hamilton, Oakville and Mississauga are legitimate destinations – but the more people we can remove from Lakeshore rails and the QEW the better so that inner GTAH growth can be accommodated.


  11. I agree with Mark that we should be looking at ferries regionally. The TTC proposal never made a whole lot of sense, but for destinations east of Hamilton there’s definitely a case to be made. Certainly St. Catharines should be workable if it has proper bus conections.

    Does anyone know how badly the last private hydrofoil operation actually did? Did they have any connections in Niagara? I don’t really know anything about them except that the service existed but didn’t last long… Has GO ever looked at it at all?

    Steve: I don’t think GO ever looked at ferries. The basic problem is that their capacity is much less than that of a train unless you are talking about a really big ferry, and when they get to Toronto, all of the passengers still have to get into the core somehow. Dumping them all on the Harbourfront car isn’t much of a connection. There are also severe problems when weather prevents ferries from sailing and the trains won’t have the capacity to absorb the overflow.


  12. If you want to look at ferries look at the total travel time and I bet that a GO train is faster. Ferries travel at about 20 to 25 mph depending on the waves. It would take 1.5 to 3 hrs to cross Lake Ontario, again depending on waves, plus the connection times from downtown to ferry terminals. You would be in Toronto on the train long before that. There is also the problem of really bad weather that would stop ferry traffic. Not to merntion the capacity problems about getting the passengers onto the 509 and 510 cars.


  13. As far as any ferry connections with appropriate capacity in an appropriately integrated transfer is concerned, I think Rouge Hill GO is the only feasible candidate. However, I’d wager the market for that is rather poor, regardless of how feasible that may be.


  14. I do believe the GO Ferry idea is worth investigating, far more than the Scarborough/Etobicoke ferry.

    Whilst the winter weather on Lake Ontario gets bad, there are many other ferry services in the world that function in as bad or worse weather – Ireland to Britain, Sweden across the Baltic,

    If the weather’s so bad, the Burlington and possibly Garden City Skyways would be probably closed as well, making this a work-at-home day.

    I disagree with Robert Wightman that a GO train would be much faster than a GO express train from St Catharines say. Even the fastest Via schedule, with minimal stops, takes 1.5 hours to travel this distance.

    Steve: The North Sea ferries you refer to are immense vessels compared to what is likely to run as a GO ferry locally on Lake Ontario. As such, they are built for much heavier weather. They require considerable terminal facilities both due to their size and due to their carrying a large complement of vehicles, not just passengers. These don’t exist in either Toronto or St. Catharines. One important point that came out in the TTC’s study is that there are speed restrictions in the Port of Toronto due to the large amount of existing boating and constrained paths for vessels. We’re not talking about something like the Vancouver Sea Bus that can just scoot across the harbour.


  15. How about a downtown relief subway straight down Bay street from Bloor to the Waterfront?

    1. It would alleviate the congestion at Yonge and Bloor.
    2. It would provide a more direct connection than even the Yonge/University Line to the waterfront, Air Canada Centre, the financial towers, city hall, the bus station, and all the condos going up on Bay north of there.
    3. The east and west waterfront LRT lines won’t have to turn up to Union Station. Their can terminate at the end of the Bay Subway, with a single connection to all downtown destinations.

    The big question is, of course, how deep would it have to go to get under the Yonge-University Line at Front Street, and would it have problems with flooding?

    Steve: The problem with a junction at Bay is that you would make the connection at the peak point of the line and would not reduce demand on BD. Passengers from the west already have the option of changing at St. George and would probably continue to do so. Those from the east might split between the two lines, but the real challenge is to drain off north-south traffic.

    Physically, getting under the existing structure at Bay and Front would be tricky, and the existing Harbourfront line’s tunnel is too small for subway cars. There are crossings of pedestrian tunnels under Bay itself to consider, and I suspect the S-curve at Queen could be a challenge for cars with subway clearance and turn constraints. Where one would place a track connection to the existing system is a mystery. If you really insisted, this would probably be better as an LRT “subway” that could be operated from the new Portlands Carhouse.

    A DRL to Eglinton and Don Mills would intercept some traffic and would provide a completely different route into downtown than simply doubling the Yonge Subway. A Bay Street line would be deserted outside of the rush hour because it would offer little more than what is there now.

    No matter what we build, it will cost a lot (including the option of trying to force everyone onto the Yonge line). The question is which option gives us the best value in improved network performance and travel options for the money invested. I know I am sounding like a flaming right-winger, but there are times it’s necessary to force a discussion that looks at the network, not just one pet project (oops, “professional proposal”).


  16. I agree with the comments that adding additional tracks at Union will be difficult. However, much of the future train capacity problems at Union can be fixed by changing the way trains are loaded, and what trains do after a “run”:

    CURRENTLY: When taking Via, everyone is queued up inside the terminal and only allowed to go through a very slow boarding procedure *after* the train has arrived. GO has many trains sitting in the station for long dwell times.

    FUTURE: It may require some minor reconfiguration of platforms, but people shoudl be allowed on the platform prior to train arrival. Dwell times kept to a minimum. *Every* train should be a through train.

    Steve: This has been a big debate with GO Transit during the Union Station project design. GO wanted to have flexibility in platform on which they placed trains. Their original scheme was that passengers would wait in the concourse until their train was listed and then go up to the appropriate platform. This turned out to be impractical because of pedestrian congestion that showed up in the modelling, although anyone with a few brain cells to rub together and a basic knowledge of the station could have told them this.

    Via wants to continue queueing passengers in the concourse.

    For GO, the source of the problem is one of passenger safety, that is of having all the moves on and off of the platform occur while trains are stationary. If dwell times are shortened, there would be moving trains while there were hundreds of pedestrians moving about nearby. I’m not sure how they plan to resolve this.

    For Via, the issue is that some trains have specific cars assigned to destinations. It would be difficult to marshall passengers onto the platform in advance to fit this arrangement. Moreover, Via passengers have luggage and take up a lot more platform space.

    As for hookups of trains east and west of the station, this is done to some extent today by GO, but there are problems in matching schedules and train lengths. As services become more frequent, I suspect that operational simplicity will force through-routing of more than the Lake Shore service. Via hooks up some trains today. For example, my return trip from Stratford, arriving just after 11 am in Toronto, became an Ottawa train. This arrangement has advantages in train and crew scheduling, not just for platform usage. However, it also requires a considerable amount of platform time.

    To me, one fascinating aspect of the entire project has been the pedestrian modelling which is quite sophisticated and which has resulted in design changes to the proposed station layout. Originally, only the AM peak (arrival) was modelled, but later the PM peak was added. This has a completely different behaviour pattern, and problems emerged that had not been evident in the AM model.


  17. A major problem for GO and Union Station through routing is that GO has three major (Lakeshore West, Milton and Georgetown) plus one minor (Barrie), while the east end has one major line (Lakeshore East) and two minor lines (Richmond Hill and Stouffville). The Lakeshore lines though route but there is nothing there for Georgetown or Milton to through route with. At one time the then 3:58 p.m. train to Brampton deadheaded back to become a Stouffville train and there are probably other examples but nothing that works as s passenger carrying schedule.

    It would probably be an advantage to GO to run the Georgetown and Milton trains through Union to a yard in the east where they could wait out there recovery time, change ends and then head back into the station to go back west. The other alterantive that is apparenly being looked at is a stub end station to the west of Union for some of the GO lines. An elevated or underground station west of York or Spadina would be nice but I would hate to think of the construction costs or the disruption. Perhaps they could put a 4 track line under Front and have 200 foot long platforms between Yonge and York. It could be part of the PATH system since they use POP. Whatever GO does it is not going to be easy or as convenient as a bigger Union Station would be. Does anyone know if GO still puts two trains on the same platform, one going east and one west, like they used to do. They had a Richmmond Hill and a Georgetown train on the same platform loading at the same time. This does improve track usage but can result in 2200 feet of train on one track and long walks.

    VIA and Amtrack have to re-think their entire passenger train philosophy and abandon their archaic loading methods and ticketing systems. They need to run double deck type equipment that has at least one double width door per car that would open at every major station. The cars should have aircraft type luggage compartments plus areas by the doors for large bags. People would load and unload faster, there would be no 20 minute platform wait while every single ticket was checked and punched or written on like they do today. In European stations a train will come in at 9:58, load and unload a few hundred people and be on its way by 10:00. It may be to much to think that VIA and Amtrack could advance to the 21st century but could they at least get into the 20th century.

    Union would have more capacity if the trains running through it were run more efficiently. Perhaps the overcrowding will force everyone to apply more modern thinking and we would all be better off. Some of it could even rub off on the TTC.

    Steve: Once the planned upgrade of Richmond Hill service occurs, at least some of the Georgetown service will have something to hook up with. As for Milton, some of those trains may wind up at North Toronto Station hooked to a North Pickering service.

    As for passenger moves, GO originally wanted to hold all departing passengers in the concourse until just before train time to give them flexibility in track assignments. Modelling of pedestrian flows in the pm peak have convinced them that this won’t work. VIA gets away with their huge queues and a ticketing system that makes the TTC look positively futuristic because there are so many fewer intercity trains than in bygone days. As VIA builds up its service level, the idea of a queue backing up into the Great Hall will simply be unacceptable. (Part of the new design deliberately avoids contention between GO and VIA passengers.)


  18. From what I’ve read are any of the trains that use Penn Station actually running on diesel – I don’t think so. It appears that all these trains use electricity from either third rail or catanary just before they enter Penn.

    “The Genesis P32AC-DM are a regional service dual mode locomotive found exclusively on almost every train between New York and Albany (These are Empire Service trains, the Maple Leaf, the Aderondeck, the Ethan Allen Express, and the Lakeshore Limited). This is beacause trains entering Penn Station can not be diesels. The trains use there diesel locomotives throughout the trip from Albany until just near Penn Station when the driver puts the third rail shoe down and uses this, there is no change felt to passengers on board. The train then stops at Penn Station, discharges all regular passengers and then runs light under the east river on the third rail, and into Amtrak’s Sunnyside Yard. Sometimes the Lakeshore Limited doesn’t have it’s P42 locomotive(s-at times it runs with two) from Chicago switched to a P32AC-DM in Albany, this means the train pulls into the tunnel just before Penn Station where the catanary as started a electric AEM-7 locomotive is attached, and the train is pulled into the station by it. Most tracks in Penn Station have both a third rail, and catanary, New Jersey Transit and other Amtrak trains use the catanary, The Long Island Rail Road, uses the third rail. Metro-North Railroad, who owns the tracks between Poughkeepsie and Spiten Dyvil, uses the same P32AC-DMs on its diesel trains for electric service into Grand Central. From the outside the P32AC-DMs look the same as the P42s, except the P32AC-DM are numbered between 700 and 717, as the P42 are numbered from 1 to 207*. From the outside the third rail shoes can also be scene.”

    Steve: Where are you quoting from?


  19. “For Via, the issue is that some trains have specific cars assigned to destinations. It would be difficult to marshall passengers onto the platform in advance to fit this arrangement. Moreover, Via passengers have luggage and take up a lot more platform space.”

    This is not a real issue. Amtrak does it at numerous stations; the car attendants stick their heads out of the cars and yell “Syracuse passengers here!”, and also ask people getting on the cars what their destinations are; they point people down the platform if they’re at the “wrong” car. If the train is in a hurry to leave, they let them on and then the passengers walk through the train (it’s not going to split in the first ten minutes). Unlike VIA, all tickets are checked *on the train*.

    On top of that, when they’re *really* on top of things, they tell people in advance where to stand on the platform. There are numbered signs at intervals along it denoting particular locations, and you can wait at the location corresponding to the car you’re gong to get on.

    Passengers are kept off the platform until 5-10 minutes before train departure time, but then they are encouraged to get on the platform — no queueing involved — so that the train can depart on time. The tickets are checked on board, of course.

    The only station where I’ve seen any queuing is Union Station Chicago (though there may be a couple of others), where they use quasi-Via-like queuing to keep people off the platforms (tickets are still checked on the train). However, that is a 100% stub end terminal station, as well as being the terminus for every train which goes there, and it probably doesn’t make that much difference there (criss-crossing throat operations are the limiting factor at that station, not platforms or passenger movement).

    Incidentally, Amtrak trains basically do have airline-style luggage compartments and extra spaces for large bags. The baggage cars are for *extra* baggage, and it is loaded in a separate operation running in parallel to the passengers (and often faster, even when the train is stopped for under a minute).

    Via Rail’s boarding practices are completely insane and are not used anywhere else in the world. While Amtrak’s practices still involve fairly long dwell times compared to commuter trains, they’re much, much quicker than Via’s, according to everyone I’ve heard comparing them.

    Regarding commuter service, an electrified Lakeshore line should require two platforms (one in each direction) permanently, and no more (four-tracking would imply NYC Subway scale traffic, which is not going to happen), and very short dwell times. With appropriate pairing and through-running, given ridership, all of the other lines should fit in two more platforms (one each way) provided the trains have clear signs on the side announcing which line they’re on. (Oops; the current ones don’t.) Surely GO has enough platform space to manage that?

    Of course, this would require significant yards both east and west of Union Station, quite near downtown, to store trains midday, akin to Sunnyside and the West Side Yard in New York City. The total elimination of reversing/terminal operations in Union Station would be necessary to get the maximum use out of the minimum number of platforms. And I just don’t know where such yards would be found, particularly on the west side. The east side looks like there are some possibilities. Luckily an east side yard is more important because the west side lines have more traffic.


Comments are closed.