Hacker News Reader

Comment by ThinkBeat | Hacker News Reader

From life in Oslo, Norway.

In so far as improving the comfort of passengers using mass transit the best way to attract more travelers and retain travelers is to run enough busses, trains, subway trains, during rush hour so that the passangers are not packed like sardines ass to dick for 30 mins or more twice a day.

Then cooling inside the inside the passaenger area is also key Sweating like a pig while playing a sardine makes the experience even worse.

Starting work drenched in sweat and the smell of other poeples perfume and cologn is not a good start.

View on HN

I agree, but that costs a lot money. It costs $100/hour (this is likely low since I haven't checked in several years what inflation has done) to run a bus. That is one bus, even a tiny city will need more than 100 buses to just provide bad service, and would need 5-10 times as many just to give good service. If running buses were free everyone would have a bus every 5 minutes, with many different routes to choose from. Providing good places to wait for your bus is cheap by contrast, the cost of a single bus will upgrade many bus stops and the investment will last longer than the bus (though you need to maintain it)

You are wrong about rush hour. You need to provide good service all day, not just during rust hour. People who have good rush hour service and then a "family emergency" pops up mid day when they discover service is bad will go back to driving just in case that happens again. Plus running all day service at the same rates ends up breaking even - between staff willing to take less per hour, and good service enticing people to ride you can generally break even on your costs.

> even a tiny city will need more than 100 buses to just provide bad service, and would need 5-10 times as many just to give good service

Wikipedia says De Lijn (the company that provides public transport for Flanders) has 2200 busses total. Service could be better but it's not awful either, and that's for a whole region of 6 million inhabitants, so I think your numbers are quite off. A tiny city only requires a handful of busses, a medium one use 50 busses or so.

I assume that the major urban areas also have decent rail and tram systems, though. It makes a big difference.

Dublin has a metro population of about 1.5 million, and has about 1,400 buses (excluding intercity and private services); however its rail and tram systems are, er, inadequate. Berlin has about 1,600 buses for a metro area of 6 million, but of course has a far better rail and tram system.

Right, I totally forgot that the bus network here is only there to fill the gaps between the tram lines, which might not be the case in other places.

Yeah, there's a route (or set of closely related routes) near my office which, at peak times, has a 100 person double decker bus about once per _minute_. Really, it should be a tram route, but that would require building a tram line, and Dublin tends to see that as an immense undertaking to be done only once a decade or so (and this decade's one is already planned out).

Have they made any efforts to make peak demand... less peaky?

That's kinda a problem with big capex transit projects: you're spending all this cash to 24/7 fix a problem that only exists for 5-10 hours/week.

This is what kills me where I live, too. I can’t understand why everyone insists on commuting exactly at the same time.

Before WFH became a thing here, I just moved my office hours one hour earlier. I went from being sardine-packed to ample seating space. Bonus points for the buses and trains running faster, so my commute was not only orders of magnitude more comfortable, it was also much shorter !

Yes, I understand not everyone can do this, but the point is for people to be spread over a wider time range. Many people don’t have children to get to school and whatnot, a sizable chunk of them could probably move their hours a bit earlier or later.

Just keep in mind that a company car is a major perk all salaried employees strive to get because that allows them to commute to work for free from all those tiny villages.

> You need to provide good service all day, not just during rust hour.

Providing good service at non-rush-hour is comparatively easy, though; it just has to be relatively frequent. Particularly for buses, the trouble at rush hour is that you can effectively hit capacity for a line; you get buses waiting for other buses to leave stops, and so on.

While true, if you have good all day service and run into that problem you have more than enough demand to build a train.

You can't just plonk trains down in the middle of a city.

Everything costs money. The cost is relatively low to the public and economic benefit from plentiful public transit. Breaking even financially would be icing on the cake. Public transit is a service, we should expect it to cost money — and not necessarily pass 100% of the cost to passengers, since it needs to be affordable to provide the full benefit to people.

Let's agree that the service should be subsidized. Let's establish that Im amenable to a free service 100% tax payer.

I am, in fact amenable to this.

How do you establish fairness though?

Best case scenario, say in the US ex-NYC, you will always have 20-40% of people needing to ride their cars because the tails of the distribution are prohibitively expensive.

Nor are these people necessarily wealthy - the economically wealthiest class of people are urban dinks.

So how to do you establish fairness to this 20-40% of people subsidizing the others?

> So how to do you establish fairness to this 20-40% of people subsidizing the others?

Well, for a start, they're not, necessarily. That thing they're driving on, it didn't magically appear, you realise. Road maintenance and construction is a _huge_ part of total transport spending.

Yes, and road maintenance is proportional to the fourth power of the axle load. Busses are typically loaded at 4-8 times the per-axle weight of passenger cars.

And speed.

So many problems would go away if we drastically lowered vehicle speeds. For starters we could drive lighter built cars.

Im in favor of raising the gas tax to fully cover road maintenance.

We're back to square one - it doesn't cover the $100/hr that a cleaner has to subsidize. Give me a proposal, or an explanation, that makes this fair and Ill agree.

Like Luxembourg, just make it free for everyone! People will start using it out of convenience. Let it be subsidized by tax money, it is fair.

Busses are a band-aid that get employed as a cure too often. Ideally, you're running trams at surface level with decently-distanced stops and biking/walking infrastructure in between. Busses require infrastructure - large roads - which discourage pedestrian activity and encourage personal vehicle use. You want people walking near businesses that they might patronize, not flying past in large vehicles that make the space uncomfortable for pedestrians.

Is there a single city in the world with trams/subways/trains/etc but no buses?

Very few, if any, but there are cities that use busses for what they're good for, and not as a bad urban planner's trams.

You need roads anyway! Most of the businesses need to get trucks in to deliver the things they need. Thus you can reuse those roads for buses for places that are less busy and this is cheaper than a train. The real trick is how to get private one passenger automobiles off the roads.

In very dense areas you should have both roads and trains. However only the most dense cities in the world can support that, and then only in their most dense which is often small. For everyone else buses save money over trains.

But buses actually don’t. you just pay for it later. Roads are one of the fastest deteriorating forms of infrastructure. the hole “ but rail dosnt make money” argument ignores the fact that… roads don’t ether

The time frame before a rail line will save money over buses is measured in decades.

With buses, you don't include the cost of road maintenance because buses share the roads with millions of other vehicles so the allocable cost is neglible.

Extept that you have to maintain the roads anyway. taking a few buses off won't save much so long as the road for trucks exists.

Roads for buses and trucks are far more expensive than for cars. Car drivers are absolutely subsidizing bus passengers by order of magnitude. Per passenger mile even Uber is cheaper.

Not when car volume is several orders of magnitude higher than bus volume. Bus passengers are subsidizing car drivers.

Busses and trucks damage roads up to 30,000 (yes 30k) more than cars. Cars absolutely subsidise all roads.

And btw neither buses nor trucks nor cars have some constitutional right of way.

Most European cities are transitioning towards dedicated bus lanes. In that case, in principle, that can be light rail or tram too. The biggest problem is crossing with normal roads. Trams are more tricky to implement than busses.

A one lane road with minimal parking can handle a crazy number of deliveries and busses. Allow cars on that road and suddenly it’s almost useless as a transportation option.

Thus the urban infrastructure dilemma.

I’m not sure it’s fair to call them a bandaid.

But fixed transportation infrastructure like trams, metro and other trains have a tendency to define and connect a city in a way that bus routes don’t.

The stops along Tokyos Yamanote line grew into major cities in their own right.

Copenhagens first metro was planned to be financed by building stops in in or underdeveloped areas and selling the land (it failed as the financial crisis of the naughts coincided with the selloff)

Bus routes have never invigorated or defined a city like fixed infrastructure. a part of a city does not become desirable just because you set up a bus stop, but a tram or metro station?

Busses are huge though. You can get fares cheaper than the cost of gasoline for an equivalent single-passenger trip just by using the available seats and ignoring standing room completely. The seats are pretty wide, so that satisfies the anti-sardine requirement easily.

The point about needing good service all day, not just during rush hours, is more interesting. If your hypothesis is that people just want the experience to not be terrible while they're in the bus, more trips during rush hours definitely suffices. You do need a minimum level of service in the middle of the day for the idea to be practical for some people. Surely 15-30min routes are fine though?

Per 15 minute is maximum for good service. 30 minutes is infrequent. 10 is very good.

7 minutes max is good. Anything longer than that is bad, though you can accept up to every 30 minutes, anything more than that is unacceptable. I used to ride a every 30 minute bus, and I always started checking where the bus was (as opposed to the schedule/time!) long before I wanted to leave. A couple times I made a mistake and arrived at the stop just after the bus left, and was then in for an unacceptably long wait for the next bus.

Mid day you need better service than rush hour. People going to work generally have a schedule and they have input into it - they can plan their day around when the bus goes. People doing errands in the mid day do not know when they will be done, if there happens to be a line when you go to the checkout you miss your bus. When you are at work you have more control over when you leave and so will always make your bus.

When people see your bus system as a way to get to work and a car is for everything else you can do worse service than when you want people to get rid of the car.

Do you have any source or more information on the $100/hr figure? That seems shockingly high to me. It only costs about twice or three times that to run a small airplane or a farming combine.

Is it surprising though?

Without looking up the data, off the top of my head.

The bus driver needs a commercial license with air brakes, is a public employee and has benefits.

He has to be making 75k/year + overhead. Let's say overhead is 50%

$56.25 / hr.

Then there's the diesel cost. A large bus must get ~ 10 mpg highway, but a city bus is all stop and go - disastrous for a vehicle as massive as a bus. Let's say it gets 4 mpg.

$4/gal diesel * 30 mph average speed / 4 mpg = $30 / hr on diesel.

We're at $86.26/hr.

You could argue my driver makes too much, and my mpg is too low, but I havent included:

- routine maintenance

- fixing broken vehicles

- cleaning

- amortization

So, play with the numbers if you will; but $100/hr has to be a good, round estimate, for running costs.

> Then there's the diesel cost. A large bus must get ~ 10 mpg highway, but a city bus is all stop and go - disastrous for a vehicle as massive as a bus. Let's say it gets 4 mpg.

Increasingly, most buses would be either electric, or heavy hybrid (ie electric with a diesel generator).

But also, consider the money they take in. My local bus system's buses take up to 100 people. The charge is 2 euro for a journey where the last leg starts within 90 minutes of the first leg, on any mode of transport. However in practice the average journey would be in the 30 min range. So if the buses are constantly full, that's 400 euro per hour takings!

Of course, it's not really that high; the buses are not always full, some people are using monthly or annual tickets, which are cheaper, some people are kids or over 65 or otherwise get cheap or free travel, the trains and trams also have to be paid for, and so on. But it's sufficient that they were actually able to cut the price (it used to be up to 3.50 for a single bus journey); the increased use from the cheaper simpler system offset the reduction in revenue per journey.

The problem with buses is that utilization is very low. Most buses, most of the time, run near empty.

This also significantly i creases their overall passenger mile GHG emissions.

Hybrid buses and CNG probably make sense. But I doubt, fully electric buses make sense: as long as dispatch-able power (ie your marginal producer) is a thermal plant you get 60ish% thermal efficiency at best minus transformer and transmission losses.

A bus engine for use in a hybrid can be made more than 40% efficient (miller cycle, high compression, high octane CNG, full throttle PWM operation, etc).

> The problem with buses is that utilization is very low. Most buses, most of the time, run near empty.

Huh. That doesn't resemble my experience, except for rural/small-town services, which are very much provided as a form of subsidy, not a viable business.

> But I doubt, fully electric buses make sense: as long as dispatch-able power (ie your marginal producer) is a thermal plant you get 60ish% thermal efficiency at best minus transformer and transmission losses.

That's still quite a bit better than diesel engines, where for realistic engines you're talking about 40% _in ideal conditions_, but much worse in stop-start conditions (you will do better with hybrids, granted). Gas also have considerably lower CO2 intensity per thermal watt than diesel does (your fossil fuel electricity production is probably mostly gas in most countries).

But, also, in countries with a lot of unreliable renewables, buses charging overnight can be used as a power sink. For instance, in Ireland, our wind generation can be anything from almost nothing to greater than total system demand; reasonably often, wind turbines actually have to be stopped to reduce output, especially at night. At that point, charging the bus is for practical purposes free, or may even have a modest negative cost (the network will sometimes pay for power consumption).

"That's still quite a bit better than diesel engines, where for realistic engines you're talking about 40% _in ideal conditions_" and so on

Which is where hybrid drive trains and PWM comes in. Run the engine at its ideal and only at its ideal with no care for the actual load requested.

Furthermore I mentioned CNG - ie methane. Methane has more energy per CO2 gram than any hydrocarbon and has an insanely high octane number -> it can run an Otto cycle at diesel compression ratios.

Run it as a miller cycle and the thermal efficiency starts to go up even highed.

Really my 40% is a low ball of what is possible in ICEs to not offend anyone's preconceptions. F1 claims to have achieved >50% thermal efficiency in a race engine but Im usually loath to mention that because of their secrecy and therefore lack of indee verification.

And no one talks that lines and transmission is lossy. No one talks that dispach-able power is often coal or oil. No one talks about the environmental and social cost of batteries.

Hybrids ameliorate (the most, btw) just about every aspect of this while minimizing external costs

> No one talks that dispach-able power is often coal or oil.

... Because they're generally not, nearly anywhere. Oil power plants are, in general, rare today nearly everywhere in the world; natural gas really did a number on them. And dispatchable power plants are almost _never_ coal; the startup lead-time is too high (hours at least). Ignoring weird stuff like grid batteries and pumped storage plants, they're nearly always gas turbines.

For short term stabilization, say within time constant of fifteen minutes, of course you use hydro or nat gas.

But the electrical industry makes very careful predictions of the next day's power consumption; that is within a time frame that coal can easily respond to.

Since coal is expensive and dirty, they end up supplying the marginal production. Ie the buses run on coal.

But why keep coal at all though? One advantage coal over has is that you can store massive amounts of energy in a pile just outside the plant. Only nuclear plants can store such massive amount of energy locally.

This storage ability is heavily used in the North East to toughen the grid in the winter where gas pipeline pressure drops due to heating demand.

As to oil, it is actually still used, albeit intermittently and then rarely [1] It is usually used in plants that are primarily non-oil burning. Again, the advantage is that oil is far easier to store than nat gas.

Note that, since oil has a lot more carbon than methane and since oil extraction, at least in the USA, is very energy intensive, oil contribution to GHG is far greater than it's 1% of the energy mix would suggest.

And its all in the margins where the EV bus gets charged.

[1] https://www.eia.gov/todayinenergy/detail.php?id=31232

> Then there's the diesel cost. A large bus must get ~ 10 mpg highway, but a city bus is all stop and go - disastrous for a vehicle as massive as a bus. Let's say it gets 4 mpg. $4/gal diesel * 30 mph average speed / 4 mpg = $30 / hr on diesel.

We've already got decent electric bus options implemented in a number of cities, and they're only going to get better. That's before you consider hooking them up to wires like a sensible transit network.

Municipal fleets often make it a point to use alternative energy with buses, like CNG before batteries were a practical option. They do it for public image, because of the urban air quality issues with old-school diesel exhaust, and for cost reasons.

A more interesting point is pavement. Public buses, garbage trucks, and schoolbuses are some of the most significant non-weather-related causes of road wear, and this is only going to get worse the heavier the axle weight if you try to run long-range battery banks. Road wear scales with axle weight raised to the fourth power. Arguably there is a case to be made for the articulated or even bi-articulated designs if it allowed you to drop axle weight, with significant benefit to the Packed Like Sardines problem. Articulated EVs have a lot more freedom to redistribute the weight and to power individual axles more intelligently.

Its not obvious to me that EV buses (or any EV vehicle, frankly) make sense. [1]

Hybrid and CNG make a lot of sense though.

[1] Run the math and hybrids come ahead of EVs in personal vehicles for GHG emissions. For buses it would be more skewed in favor of hybrids.

In the US every transit agency published their costs to the FTA. I looked that data up about 5 years ago, and I can't remember how to search it, but it should be online if you want to figure it out. (unless Trump has removed that?)

Honestly, for 'emergency' things a taxi service is a good complement. Busses are good at regular routes between frequent routes/hubs, not for 'I need point to point immediatly now'.

The goal should be more to cover peoples normal transit needs. So plenty of routes with stops very near most people's homes and destinations, and a high enough frequency at most hours of the day and night to not cause too much waiting frustration.

> "That is one bus, even a tiny city will need more than 100 buses to just provide bad service, and would need 5-10 times as many just to give good service"

You say that as if it's an unthinkable impossibility. As of 2017 London had 8,600 busses carrying 6.5M passengers every day. Yes a big city needs a lot of buses, so they have a lot of buses. No problem. What would it cost if there were no buses and 6.5M more car journeys every day? And the roads and parking spaces for - what - 1M+ more cars to account for that?

What financial cost to the city for roads and infrastructure and traffic lights and signs and maintenance, what financial cost to the people needing cars, what time cost of humans stuck in traffic, what health cost of less exercise, what health cost of pollution, what accident cost?

>You are wrong about rush hour. You need to provide good service all day, not just during rust hour.

Your first half talks about how high the hourly rate to run a bus is. Then in the second half you recommend the same service level for busy versus non-busy times? Seems sub-optimal.

All day service ends up increasing all hours use - including rush hour.

however no matter what it is expensive and you need years of service before people change habbits. Often the expense is so high that 'you cannot get there from here'

Yes, it costs, and yes, to make public transit a valid alternative service needs to be frequent all day.

However, all of the maintenance around allowing private transport is also a cost. That includes all of the private costs for car ownership. If public transport is a valid option, then car ownership becomes a choice.

A tiny city might need 100 busses, but how many cars does that replace? What is the full cost of those cars? If everyone made a monthly "bus payment" instead of a "car payment" (actually payments when you add insurance, maintenance, ...), I think we'd find the bus system much less expensive.

  >  It costs $100/hour to run a bus

That sounds pretty cheap!

Googling suggests you can get 40 people on the average bus, but let's say that's 20! Let's be super rough to get guesstimates. California has an average gas price of $4.83/gal[0] and average car is like 24mpg. We'll take that at 60mph and assume you're driving that (mpg is usually worse in cities). So let's say ($4.8/g) * (1/24 g/m) * (60 m/h) = 12. So at half capacity the bus is more than half the price. That's even before we include things like insurance, maintenance, and anything else (like parking fees). Sure, we're super rough, but given how big the difference is I'd be surprised if refining switched the conclusion.

Of course, there's extra convenience of cars, but a lot of time we don't need that extra convenience. A lot of times we're going the same way everyone else is and our convenience becomes inconvenient. There's also lots of extra conveniences of public transportation too. If you've experienced good public transit you probably prefer it in most situations.

I agree, things need to run all day but I disagree that it costs a lot of money. I think the bigger issue is the negative feedback loop. Public transportation sucks, so no one uses it. -> Public transportation can't get more funding because no one uses it. -> No one uses public transportation because it sucks. Unfortunately this is one of those things where "build it and they will come". It won't happen right away, momentum is a bitch, but it can't ever happen without making the investment. And clearly it is actually cheap.

[0] https://gasprices.aaa.com/?state=CA

if you can get the money to buy and run all those buses and run them for several years while people figure out they are good. They also need to be good - many buses run terrible routes that won't get you where you want to go in a reasonable anount of time.

I recognize this. Please read my last paragraph.

> People who have good rush hour service and then a "family emergency" pops up mid day when they discover service is bad will go back to driving just in case that happens again

I don't know why so many people seem to think it's one mode of transportation or another? For regular commutes, take public transportation. For emergencies or when time is of the essence, use a car?

I disagree on all day frequent service. You need taxis or car-share services for emergencies. This is kind of the norm in most countries with good public transport and it just works, as long as the public transport is reliable. If it starts to get unreliable there's a downwards spiral in usage and utilitization.

100% this - make it more comfortable and handle the rush.

I live in London, we have tons of public transport.

In the middle of the day, outside of rush hour, I use it a lot and it can be really pleasant, particularly suburban rail, you can stretch out a bit, there is often air conditioning, a nice view from the window, etc.

When it's busy I will drive even if it takes longer because my car is guaranteed to have personal space and is clean.

Reply on news.ycombinator.com