This is the first time I have found myself with a question after reading the core rules, gone to the forums, and found out Alex has not already answered it.
It was weird. So I derived my own answer, using this thread (http://autarch.co/forums/ask-autarchs/constructionlabor-costs) plus the books as my starting points.
The question I’m attempting to answer is; how much can a high-level wizard help in constructing buildings, strongholds, and the like, and what should the rules for that be?
I found it to be an extremely difficult task at first because the book does not split between materials and labor. Then I realized; what if you just aren’t paying for materials?
That is, your labor costs include the cost of your men quarrying stone and harvesting lumber, then dragging those materials to the site, then building out of them. Under this assumption, having N GP worth of materials at the site simply reduces your construction cost (and time) by N GP. I use that as my core assumption for all of the following. Now on to the actual house rules!
General rules: In no case can the cost of a structure be reduced by more than 50% by magic, unless the magic being used can replicate the entire structure. For example, the cost of a wooden palisade can be reduced by 100% with a Wall of Wood spell, because a Wall of Wood is capable of creating a wooden palisade. Similarly, a castle wall can have its cost reduced by 100% by repeated castings of Wall of Stone. However, a tower cannot have its cost reduced by more than 50% by castings of Wall of Stone, no matter how many times you cast it, because the Wall of Stone spell cannot create wooden doors, floors, stairs, furnishings, and the like.
A spellcaster who is assisting in construction must work closely with the artisans and engineers. The necessary castings will be spaced out over the course of construction; thus a wizard assisting in the building of a small tower (5 walls of stone being cast over 28 days, the derivation of which will be shown later) would have to cast Wall of Stone the first day of construction and every five days thereafter. The spellcaster can do other things during the days they are not spellcasting, but must be available to the artisans for at least occasional consultation at all times during the process.
In any case where sufficient castings are required (Judge’s discretion how many castings is sufficient; my gut ruling would probably be 5 as a default, but you could go with as few as 2), one engineer per 100,000 GP base price is required. If a qualified engineer is not hired, the structure will be unstable. Whether it collapses on its own or waits for a stimulus (earthquake, catapult, gust of wind, light tap, whatever) is left up to the Judge, but the summary is that if you want it to stay standing for more than a week or two, you should really consult an engineer. (Note that an engineer hired to work on a project of this nature will charge you for the full month even if you complete the construction in less than one month.)
A Wall of Stone spell reduces the cost of the structure by .25 GP per cubic foot (250 gp per casting, to a maximum of the structure’s actual cubic footage of stone). This number was derived from the cost for 20’ and 30’ high curtain walls, both of which cost .25 GP per cubic foot.
A Wall of Wood spell reduces the cost of the structure by 1.5 GP per cubic foot (150 GP per casting, to a maximum of the structure’s actual cubic footage of wood). This number was derived from the cost for a wooden palisade, which costs 1.5 GP per cubic foot.
A Move Earth spell can be used both for excavating and for earthen constructions. When used for excavation, it reduces the cost by .02 GP per cubic foot excavated (derived from the cost for pits and moats). When used for construction, it reduces the cost by .167 (I haven’t decided where exactly to round that yet) GP per cubic foot, derived from the cost for an earthen rampart.
Transmute Rock to Mud (or vice versa) allows you to move mud around and build with mud, then turn it into stone. This allows you to build stone structures using the cost for an earthen rampart, which saves you .083 GP per cubic foot. (The total savings for a 20’ high 100’ long 10’ thick stone wall would be 1660 GP, requiring only a single casting of the appropriate Transmute. Note that a transmuted structure cannot replace the entire structure, as you cannot simply stack up mud 20’ high, and thus must obey the maximum 50% reduction limit, if somehow that manages to come up.)
A 20’ diameter, 30’ high stone tower (small stone tower) has a cubic footage of stone of (pi10030)-(pi56.2530) = 4123.etc. (Volume of the outer cylinder minus volume of the inner cylinder, with a 5’ thick wall, gives us radii of 10’ for the outer and 7.5 feet for the inner.) This will require 5 castings of wall of stone to create all of the stone necessary for the form of the tower. However, there are only 4123 cubic feet of stone to replace, so instead of the full potential 1250 gp for 5 castings, the cost of the tower is reduced by 1030 gp, from 15,000 to 13,970. It will reduce the required construction time from 30 days to 28 days.
Note: Original math is for a tower with a 20’ outer diameter, which is a little ridiculous. With a 20’ inner diameter, the outer diameter is 30’ (5’ of wall on each side), giving us radii of 15’ and 10’. This gives us 11, 780 cubic feet of stone in the tower; 12 casts of wall of stone will reduce the cost by 2,945 GP. Reducing the cost by 2,945 GP will reduce the construction time from 30 days to 25 days.
A generic stone building (20’ high, 30’ square, 1’ thick walls) has 2400 cubic feet of stone (20 * 120 * 1). This will require 3 castings of wall of stone to fully replace, which will reduce the cost of the building by 600 GP, from 3,000 to 2,400 GP. The total construction time will be reduced from 6 days to 5 days.
A generic wooden building (20’ high, 30’ square, 1" thick walls) has 200 cubic feet of wood. This will require 2 castings of wall of wood to fully replace, which will reduce the cost of the building by 300 GP, from 1,500 to 1,200 GP. The total construction time will not change, as the price was reduced by less than 500 GP.
A stone castle wall (60’ high, 100’ long, 10’ thick) has 60,000 cubic feet of stone. Sixty castings of wall of stone can create this entire wall out of nothing, reducing the price to 0. A 9th level wizard who can cast only one wall of stone per day would take 60 days to construct this wall, which is actually slower than the normal construction time of 45 days. (If the wizard wants to just cast wall of stone as fast as he can while the workers also work on it at a rate of 500 gp/day, at a rate of one wall of stone per day, the wall will be done in 26 days at a cost of 13,000 GP. I rounded a little to solve that.)
A wooden longhouse (15’ high, 30’ long, 15’ wide, 1" thick walls) has 562.5 cubic feet of wood. Six castings of wall of wood will replace all of this. The cubic footage reduction mathematically would reduce the cost by 843.75 GP; however, this is more than 50% of the cost of the structure, and the entire structure cannot be created out of walls of wood. Therefore, six castings reduces the cost of the structure by 150 GP, or 25 GP per casting.
The summary of these derivations is that it is really not worth the effort to hire wizards to build your stronghold for you; you get a bit of a time break but mostly it’s not a huge one. I’m happy about that, because if it were economically feasible to hire wizards to build strongholds, then more people would do it. Since it is economically infeasible, that means that this allows wizards to use magic to ease the burden on their own finances if they are willing to spend their own time, but non-wizards don’t feel that they need to hire one or more wizards to cast wall of stone all day.
I’m surprised by how effective Transmute Rock to Mud turned out, though I suppose I shouldn’t be; the volume of effect of Transmute is 30 times that of Wall of Stone. It’s also much less useful at sudden battlefield walls, so it’s not really a problem that it’s better at constructing permanent walls when you hire 3,000 men to drag mud around.