News:

SimuTranslator
Make Simutrans speak your language.

Costs of tunnels

Started by AP, February 05, 2012, 05:46:39 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

AP

All good points.
Quote from: sdog on February 05, 2012, 05:41:13 PM
There should be two massive cost drops in tunnel building, introduction of dynamite (uk patent 7 may 1867) and recent prefabricated tubbings and automated drill machines.
Sounds good. Maybe we need different types of tunnels, like track, becoming available.

I recall seeing someone improving the entrance portal graphics (the current ones have a roof rather than a cutting) - presumably that's a pak issue though.

Mod note - split from the balancing discussion

jamespetts

That's very interesting. Unfortunately, I suspect that it would be extremely difficult for the game to calculate which tiles are under flat land and which under steep mountains. As to tunnelling costs reducing - are they not possibly offset by increased labour costs? Or is a million pounds in the 1870s/1880s for a 4 mile tunnel (the Severn) greater proportionately than tunnels cost to build now, in real terms?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

omikron

The 5,3 km long Gravhalstunnelen in on the Bergen Railway in Norway cost only 3 million Norwegian kroner to build in 1906. That would be about 150,000 GBP at the time. Just to show what granite underground can help....

omikron

jamespetts

Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

"As to tunnelling costs reducing - are they not possibly offset by increased labour costs?"

likely not, but it won't be possible to see, tunnels were hardly dug in anything but the best geological situations, this has changed. The Thames tunnel of the 19th century is an example for exceptions from this.

There was lower labour cost, but higher capital cost, it was a high risk operation, many tunnels caved in with complete loss of investment.

omikron

Well, another Norwegian rail tunnel, built in 1997 was 14,5 km long and cost 13 Billion Norwegian kroner (= 1,5 Billion GBP), because it was built under a lake, which leaked into the tunnel and made the building costs explode....

I don't think any general rule can actually be defined. MAybe one could make a price range, with the actual costs being randomly assigned?

omikron

sdog

possibly bancrupting your company?

to quote another game related to dig underground: "fun!"

jamespetts

Hmm, I don't think that making the costs of tunnelling random is a good idea.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

omikron

I didn't mean random, but a random price within a range, so that (example): 1 tile tunnel costs 5-15, with 50% probability that the costs stay in the range between 9 and 11, 35 % costing 7-8 or12-13 and 15 % 5-6 or 14-15. Do you get my idea?

omikron

jamespetts

Yes, I understood that it was within a range, but I think that there are big problems with having even constrained random costings: this would be very confusing for players.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

omikron

OK. I thought experimental was meant for people who are not getting easily confused :-)

And building a tunnel is always risky, so why not?

omikron

AP

Quote from: omikron on February 05, 2012, 08:54:57 PMOK. I thought experimental was meant for people who are not getting easily confused :-)
He has a point!

I quite like the idea of random-within-a-range. It would mean that short tunnels had a small variance in price possible, long tunnels a much larger one. It would mean small companies would only prudently undertake small scale works, which is realistic, whereas large companies could commit to larger works. Real companies can have/have had financial difficulties due to works costing more than predicted, factoring risk into the game for large structures would encourage "responsible" gameplay, perhaps?

jamespetts

I think that even people who are not easily confused would be confused by random costs! It would make no sense to have tunnels and tunnels alone as the things were subject to cost uncertainty: that would be arbitrary. In reality, any sort of project can cost more than anticipated. For this to make any sense, there would have to be a whole system where the cost of projects was uncertain; that would be a large undertaking, and is not currently a priority.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

if it's not completely getting out of hand, a rough estimate of the building cost often is not so bad today. That's getting far out of the scope of the simulation i think. One has to live with the construction cost of tunnels not being very well simulated in the game, a straight cost increase might be more than enough.

If you really want to do something James, you could make a kind of restricted tunnel, similar to what we had before ctrl-click tunneling and underground mode. So it only could be built straight. Perhaps even build a lenght restriction.

With way restrictions you could still allow tube tunnels to be built unrestrictedly. (most likely they were channeld and covered anyway?) At some point in time large tunnels would also loose the restrictions.

If one really would get into the matter, a geological map could be generated at start of the map. The height structure is then based on this geological structure, drainage decided, major rivers laid, terrain eroded and basins filled up with silt. Based on this restrictions on tunnel building and cost calculations could be done. i doubt it would be worth the whole mess.

wlindley

How about making the cost per unit of tunnel inversely proportional to its altitude?  Tunnels at -1 (under sea level)) could cost much more than at 0 or 1, and progressively less as you rise. 

AP

Would that take into accound mapsettings, e.g. I often play with sea level at -6 or-7, to get more land and less ocean.

jamespetts

It would be possible to make cost relative to sea level, certainly, but the issue is: does tunnelling actually work like this? Does tunnelling on a high plateau really cost that much less than tunnelling on a low plateau?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

ӔO

#17
tunnels cost more, or are impossible with technology at the time, on certain sub-surfaces.
I believe that the Chicago L is mostly elevated, because the ground was unsuitable for tunneling.

You might want to look on the wiki page of tunnels, particularly, 3.6, 3.11 and 3.12

There was also a documentary from discovery on the channel tunnel that might be interesting to look at. I believe it was from the series Mega Structures (Super Structures).
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

el_slapper

Hi everyone, my first post over there. Though a long-time player & lurker.

That idea of variability in building(here, tunnels) costs is really good in terms of realism. However, I fear it's a can of worms waiting to be opened.

Anything built is dependant on local constraints. Roads in the desert or in the jungle tend to cost more to maintain. Randomness still exists in building projects, with more variability for bigger projects. Mainly because bigger projects are rare. If you build 10,000 times the same house, the same road, the same railroad crossing, costs are well known & mastered. But for huge projects, well, you are basically doing a new design each time. And, therefore, there are many unknowns. I'd say that due to the tricky & changing state of the soil, tunnels are especially prone to that kind of effects. But if you begin with tunnels, you'll have to code - and balance - a biiiig number of variations.

I don't say it shouldn't be done; just, think strong before beginning. A proper balance of variable building costs can be an absolute pain to achieve.

sdog

hi el_slapper, welcome to the forum, as a participant.

I'm not so sure if a height level dependency of tunnel building costs is sensible at all. it seems rather arbitrary to me.

omikron

If you want to make tunnel costs less linear and more like reality, then some arbitrqariness cannot be avoided. As said above by someone, every major tunnel project (and to a lesser extent bridge, track and road project) is hugely dependent on a variety of factors which differ hugely from case to case, so the same tunnel length in two different contexts may cost up to ten times the same amount, if not more.

My take on this would be to make random costs within a range - the range need not be very large, but it makes things a bit less certain. How many of you calculate the cost of a tunnel manually before building it in simutrans anyway? If you do it with the key combination (which I forgot), then the result being a range would not distrube you very much, would it?

I'm simply asking, but I will give up if there is massive opposition...

omikron

AP

Maybe it is a feature which could be introduced later, once the game is more well-balanced and stable. That would make James' life rather easier I'm sure!

sdog

it would be a bit difficult to have a random value for the whole tunnel.

If you do it straight forward on a per tunnel tile base, this would only matter for the shortest tunnels. When building longer tunnels you would rather quickly get close to the expectation value.

A more general remark, one can be quite a bit arbitrary, in finding good and useable approximations. However if it is very uninformed, or based on a very rough model, as we have here it is better to sticky to simple models. Above Omikron described how different the costs for a simple tunnel in hard rock are compared to james' examples in difficult terrain. We flatten that much larger effect, but should introduce a much weaker effect addressing the uncertainty of a tunnel project? In such a case, better get a geological model for map generation. (which is entirely possible to do, has quite some benefits for industrie placement, just quite a bit overkill)


In simulations one has to look what effects are important, and what less so. There's quite a lot much more urgent in the economic model. Building tunnels is not a very important aspect of the game.

AP

Quote from: sdog on February 06, 2012, 09:10:29 PMThere's quite a lot much more urgent in the economic model. Building tunnels is not a very important aspect of the game.
I agree with the first statement, disagree with the second.

There are a lot more important elements needing balancing.

However, as the recent online game showed, later in the game players are very happy to build significant tunnels (under whole towns) where in real life this is unlikely to have occured. That this is done suggests that there is no adversity to the idea of building miles of tunnels - which should be considered a considerable undertaking done only when there is no reasonable alternative.

sdog

Making it terribly expensive should do the trick.

The rest is a kind of players code of conduct not to play too abusively. Rich players can terraform the whole map, delete towns and build a railway in the sky with viaducts. As long as there's an abundance of cash there's no way to stop. Trying to build artificial restrictions often just causes a financial broker effect. Completely reasonable bank managers invested in terrribly unreasonable ways, just because they found a loophole in the strict network of regulations. The legislative and executive reacted, closed the loopholes and the game really started.

jamespetts

El_slapper,

welcome to the forums! You make a very sensible point, in that variable building costs are a large and very difficult area, not ready for consideration until such time as we are rather more advanced in the more basic echelons of game balance, which, if executed correctly, would hopefully prevent players from having such enormous amounts of money that they can afford to build vast tracts of railway network underground just because to do so is more convenient.

The priority for this year's development is balancing, and making tunnels expensive enough is part of that work. Thank you all for your input into this topic!
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

ӔO

Quote from: AP on February 06, 2012, 09:52:50 PM
I agree with the first statement, disagree with the second.

There are a lot more important elements needing balancing.

However, as the recent online game showed, later in the game players are very happy to build significant tunnels (under whole towns) where in real life this is unlikely to have occured. That this is done suggests that there is no adversity to the idea of building miles of tunnels - which should be considered a considerable undertaking done only when there is no reasonable alternative.

part of the problem is in how you have to connect the city halls to get any passenger flow, but the cost of demolishing buildings to build your station next to the city hall is astronomical. And the horse carriages are very inadequate to making this connection. Not only that, but cities can't grow beyond double tracks without some luck and major stations require a lot of room. Even the larger cities would have problems if you try and fit 2 or 3 major stations from different players.

Around the 1850's the tunnel upkeep was far beyond sustainable, but by 1880's, the profits from expresses allowed a lot of the upkeep to be negated. It's not so much the expense, but the upkeep that held back tunnels earlier in the game.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

merry

hi,

re: tunnels & underground stations, and indeed urban railways, the national preferences & circumstances seem to have heavily informed choices IRL.
Example:
(1) In the UK, we just don't do large underground stations. Never really have. But we seem to be very happy to put railways on viaducts through our cities. And indeed, many central city stations are largely on viaduct (or in cutting).
(2) Spain: Barcelona & Madrid both have large central stations wholly or partly underground. Barcelona in particular - it's huge! But I didn't see many viaducts even thought the approaches are often from higher ground, indeed many approaches were underground.
(3) Paris (also Holland & Belgium I think) have both put suburban stations wholly under the associated mainline station, to allow expansion of capacity without enlarged footprint, and avoiding complex approach trackwork.

i would suggest trying to encourage one particular approach by financial means is not necessary nor appropriate, provided it is overall balanced...ground level city track needs to buy costly property, elevated track needs expensive viaducts, and it's not cheap to tunnel. But all are valid solutions and frankly all will be done - to large sizes - if the money is there.

re: variable building costs, even I think it's is a step too far, and will make the game pretty unplayable to all but the most dedicated - unless you also implement gradual (timed, not instant) construction and gradual cost billing, allowing a player to abandon a works that is turning out too costly. Set a budget, report back on progress, etc. A bit micro-management oriented for many players.

re: terraforming, I think the costs are already high, and limit the poor player's ability to do works in their early build. But as cash builds up, the ability to build smoother routes improves. This is normal. A rich player will be able to do anything permitted by the civil authorities.

whether all these are desirable situations is questionable, I don't do network play (way too intermittent & short time available) so can't comment there.

merry.

AP

Quote from: AEO on February 07, 2012, 01:16:25 AMpart of the problem is in how you have to connect the city halls to get any passenger flow
Since when? Can someone confirm? I thought the city halls were just an interface...

ӔO

Quote from: AP on February 07, 2012, 07:48:11 PM
Since when? Can someone confirm? I thought the city halls were just an interface...

this was always the case.
for any significant flow of traffic to be achieved, the city halls must be connected. Having the station within the city boundary is not enough.
My Sketchup open project sources
various projects rolled up: http://dl.dropbox.com/u/17111233/Roll_up.rar

Colour safe chart:

AP

Well we learn something every day!  ;D

sdog

the first 250 (?) inhabitants live in the city hall. thats also the reason why very small villages get created only with cityhalls but no residential buildings.

jamespetts

I wonder whether there is a case for reducing the number of passengers generated by a city hall?
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

sdog

this would also make hamlets to sprawl. (just imagine city hall as a more densely built village core, afterall it's on an area of a quarter square kilometre! (4 in standard)

You could increase building population and reduce passenger factor. This would tone down the effect of the town-hall. It is however, i think, not undesirable. It is the most likely place one would travel to and from in a city. in a way you could see it as the place where all the folks from sourounding farmsteads go to pick up their transport.

el_slapper

Quote from: merry on February 07, 2012, 12:53:28 PM(3) Paris (also Holland & Belgium I think) have both put suburban stations wholly under the associated mainline station, to allow expansion of capacity without enlarged footprint, and avoiding complex approach trackwork.

+1, though this is mainly true for Paris. Other french towns mostly have surface elements(with a few exceptions).

I guess the sheer size of Paris did push towards different solutions. There's another point : before the 1936 nationalization, there was 3 private companies building subways & fighting each other. It has several consequences, still today : an absolutely excellent subway coverage, but, at the same time, a , hum, subpar optimization of the interconnexions. BIG underground stations have been built later, once everything went state-owned, in an effort to reduce the chaos of the network. I'd say that effort was partly successful.

Another consequence is that Paris's underground is an absolute Swiss Cheese, and the relatively new RER E line is built one level under other lines. Stairs down are endless to get there. In St-Lazare(one of the stations in Paris intra-muros)I need nearly 10 minutes of walk from the RER E to my surface train(bringing me back home in my suburb). But I'm clueless on how to translate those difficulties in terms of gameplay.