1. Find the right spots where the soil is suited for sustained farming. These spots are scattered around egypt the way marble deposits are, and finding them should be of a difficulty comparable to marble prospecting or dowsing for ore. Each spot is suited for a particular kind of vegetable.
2. Build pumping stations on the edge of nearby water. The stations work like mini Deep Well Mines, using a single screwgear to move water into a reservoir at their top. The cost of a pumping station depends on its height and the size of its reservoir. There's a minimum distance between pumping stations, like with water mines.
3. Build cut stone aqueducts to bring the water to the fertile spots. Aqueducts come in pieces which can be straight, or corners, or Y or T junctions. They have to fit together so some kind of auto-aligning (like with brick rack alignment) will be needed. Each piece rests on a kind of pillar which is used to adjust its height. The cost of a piece depends on the size of its pillar.
4. Add leather waterspouts to the aqueduct pieces which are near the fertile spots. These spouts will tap water from the aqueduct and sprinkle it evenly over the surrounding terrain. Some sort of leather, oil, and iron bar arrangement seems appropriate.
5. Prepare the fertile spots by laying down a bed of dirt, saltpeter and whatever else seems reasonable, and seeding it with the appropriate vegetable seeds.
6. Maintain the irrigation system by regularly winding up the pumping stations so that their reservoirs don't run dry. You can do this yourself, or get people to help (possibly for a share of the veggies).
Aqueduct design
To get the proper flow of water, aqueduct pieces are slanted so that the side where water flows in is 1 foot higher than the side where water flows out. With junction pieces, the difference is 2 feet. The pieces have to match up in height as well as location. The first "piece" is the reservoir of a pumping station.
Terrain counts, so a 5-foot-tall piece built at 10 feet elevation will have water flowing in at 15 feet and water flowing out at 14 feet. If the next piece rests at 5 feet elevation, it will have to be 9 feet tall. This can make hills both a curse and a blessing -- you have to build tall enough to get over them, but they do make the individual pieces cheaper to build.
Water management
Every game hour, a reservoir will use one unit of water for each endpoint of its aqueduct system, and one unit of water for each waterspout attached to its system. (And endpoint with a waterspout counts as just one, though.)
As long as the vegetable beds get water, they will grow vegetables which can be harvested, a bit like the way a greenhouse works. The rate of growth will be slow: enough to be useful, but not enough to replace ordinary vegetable growing entirely. You'll have to harvest several times per RL day to get the full benefit.
If a reservoir runs dry, the beds attached to it begin to die. Once they're dead, they're gone, though the fertile spots themselves remain and can be refertilized and reseeded. Water can be pumped into the reservoir by winding up the pumps, the same way a Deep Well Mine pumps petroleum. How often this needs to be done depends on the size of the reservoir, but once per RL day should be enough for a smallish irrigation system.
Unlike with normal vegetable growing, these beds produce only vegetables and no extra seeds. If a bed dies, the seeds in it are lost.
Passing the test
I propose a simple system where an architect accumulates a point for every bed that gets water, every game hour. This puts the focus on the vegetable beds (with aqueduct construction being an overhead cost that needs to be minimized), and on maintaining them consistently. This maintenance can be either from raw effort, or simply from the beds being useful enough that everyone helps out with the pumping. Either is a good reason to reward an architect.
Test passing should be done like obelisks: the first architect who gets 5000 points will pass (that's 7 beds for about a game month), and after that you need 1/7th more points than the previous passing score. For further refinement, these points can be regional.
Test strategies
Since the aqueducts don't count for points directly, it's best to find fertile spots that are already near water. On the other hand, building and maintaining one big pumping station is easier than building many small ones. Architects would do well to add on to existing irrigation networks rather than starting their own.
One possible "gaming" strategy is for a successful architect to let all his or her vegetable beds die, so that a guildmate can then replant them and start collecting points. This is not necessarily bad (it encourages team play, and makes sure someone stays interested in maintaining the irrigation system), but it means that preparing a bed shouldn't be too cheap.
Deep Well as possible water source.
"Rather than depending on existing bodies of water, why not use a Deep Well Mine as the source of water for the aquaduct? You could place the DWM on a modest hill somewhere out in the desert, and have your Oasis and aqueducts as a natural consequence."
Height of reservoir and its impacts.
"Something else worth considering for this is scaling the pumping difficulty with the height of the initial reservoir, and allowing multiple sizes on that reservoir. In real life, the difficulty of lifting water (or anything else) is directly proportional to how high you lift it, and a reasonable concept for this test might be to parallel that. If you choose to place your initial reservoir 400 feet in the air, you can definitely get water to flow an enormous distance away, but it'd be a lot of work to keep it full. On the other hand, if it's just a barrel on a pole next to the Nile, you can fill it very easily, but it's not going to flow all that far or last all that long between pumpings. This adds to the 'multiple methods' concept, as you might be able to be successful with a series of small, manageable things right next to a river that you can keep track of yourself but has to be micro-managed, or with a single large and very expensive construct that you organize with a group and that requires only occasional but intense maintenance."
More than just crop irrigation!
"How about tying aqueducts into canopies (or whatever the new conglomerate buildings are being called)? Certain sub-buildings might require that an aqueduct be attached to a canopy. This could provide more of a reason for people to build towns or cities--individual houses/canopies would be able to hook up to the local water grid."
"Another thing I like about this idea is that it provides several possibilities for new techs and craftable items. I can imagine Hydrology, Aquaduct Construction, and Pumping techs, and the ability to craft copper pipe, valves, etc.
I think that you could make a plausible water system with techs and materials that are available early in the telling such as basic metals and stone, so we don't need to wait for gold, red sand, raeli tiles, etc to be discovered or researched.
Smaller sections of the Aqueduct could be built with bricks (would glazed bricks be more impervious?). Also, bricks with various colors and finishes could be used to create a surface layer that had artistic merit on the larger sections of cut stone aqueducts, especially in city areas.
Being able to add water faucets around camp would add lots of flexibility in camp layout. This may be an important facet of city development around the chariot stops, since in a populated urban area, not everyone will be able to get a spot next to the river or pond."
Point Systems (few ideas here)
"Here is an alternate idea for figuring passing of the test. You get points based on how many people take water from the system, and how much. To prevent gaming, the points you get for "how much" is on a logarithmic scale, like the scores you get for funerary temple for a single tile color.
Or you could do a combination of this and veggie growing."
"I think that having it set up this way would be very difficult to "game" as the aquaduct is "owned" and cannot gain points for anyone else, and any new aquaduct while it could supply the same area would have to start at 0 points and work its way up. Things that are multiple use (obvious ones - pumping station for water jugs) would get significantly decreasing points for use in 1 day."
1/7th point increase, problems and solutions?
"If we do as suggested, and inrease by adding 1/7 with each pass, and start at a passing bar of 5000, by the time 100 people pass, the bar would be 3,148,942,294. Or to put it another way, if the first passing is 7 beds for a game month, the 100th passing would have to be 4,408,519 beds for a game month (if beds are one square coord, that's roughly 3.3% of all the land in Egypt covered by those beds -- if the previous test passers kept theirs, roughly 1/3 of Egypt would be covered just by the accumulated beds at this point, to say nothing of the aqueducts to feed them)."
"the first is 5000pts, just add 1000 (random number here) for each pass. This would keep the exponential from getting out of hand, and likely keep the ceiling in line with growing technology and need for water. Or looking just at vegetation still keeps things within reason likely."
"We could have the passing scores reset each game season or year, to get an effect similar to releasing a new obelisk type."
Assuming that most sectional Aqueducts will be built by multiple players, there needs to be some means of allocating points. One means would be to randomly select sections that had flowing water, hourly or as needed to get the totals the right size, and credit the original builders (no Aquaduct Guilds passing around section ownership) with a point for that time period. The "final brick" in a section could be placed by a deserving helper, who would become the original builder, as is currently the case with Megalopolis sections.
-add edit rehash as anyone sees fit, sorry if I missed any ideas out there on the thread-
How about adding small pump/lift stations along the way to enable add-ons for other Architects?