I never considered compost in the model, thx for the input.

The garden is too close to the house to try bull kaka, but maybe if I ever move. I have heard of aquaponics, but I have two reservations about aquaponics-not enough land [unless we chop down trees, and still might freeze]-and its not vegetarian.

thx also for the tip of compacted gravel.

My focus in writing this was mostly geared towards 1-2 large greenhouse[s] with polyethylene covering. I have considered using plastic to cover underground emc, seems I forgot to put it in writing. I did say pvc could last hundreds of years underground. Perhaps the reason I didn't elaborate more on pvc joining another material [other than maybe being sleepy that night] was that given our winds, concrete made more sense for securing it to the ground. Also the horizontal load of PVC was only 28 lbs [i mistakenly just said the max load was 28lbs].

There is an 2 hour video from a Minnesota university. In it one of the professors was growing crops in a greenhouse that was attached to his home-it also discussed other issues such as building angles and climate batteries.

.
source:University of Minnesota Regional Sustainable Development Partnerships

One of the people from epic gardening cautioned about growing greenhouses attached to the home because of humidity issues. I am not sure if that would imply material [like wood] rot, rodents/insects, mold, or other issues.

For my Musa Ice cream [banana] I do use a more expensive LED grow light rather than just an ordinary LED. But for my [normal] seedlings that graduated out of my germination station they are presently getting by on a 9w bulb [at this time] in close proximity in a storage room. In academics, a horticulturalist [which I am not] would be interested in the photosynthetic photon flux density (ppf) for the adequate lighting for a plant. A plant needs so many photons per day for a certain area, so there is a balancing by multiplying the hours lit by the ppfd and by .0036 (or 3600 mol x sec/(1000000 x umol x hour))to get a recommended daily light integral [dli]-which the ppfd can get complicated because it varies by distance from the light source. See https://bioslighting.com/led-grow-light-distance-chart/grow-lighting/ And that is the model that I should have used, but I chose a questionable short cut method commonly seen while noting it wasn't technically right-the point I wanted to emphasize was the expense per row was astronomical.
I am not going to say definitively how much better an led grow light is than an ordinary led light. There is one site (trying to sell grow lights as an amazon affiliate) that [without showing the wattage] compares a 36w grow light led to an ordinary 8.6w led, the ppfd is 401umol/sec on the former, and 27 on the latter. He claims to be making the measurements about a foot away, but it is unclear if it is a fixed distance. But assuming everything, but the wattage, is kosher, that would suggest a Sansi grow light is 4x better in ppfd per watt than an ordinary Ikea LED.

source: www.ourhouseplants.com

source: ebay

So let's choose say a Musa acuminata, and https://growlightmeter.com/light-requirements-for-plants/ suggest the DLI is 4-14, and the recommended ppfd is 100 – 350umol/(m x m xs). Already his sansi light is too close, additionally circles may not be ideal geometrically speaking-but for convenience let's say it was a 36w t5 growlight and at 18 inches it was 350umol/mms. it means the light would have to be on 3.7-11.1 hours a day. If in contrast we used 5 ikea 8.6 watt bulbs to make our 3k lumens, kept them at our 1' position to maintain 135umol/mms, it would need to run at least 8.23 hours. Or if we used our original 4, for 108 umol/mms it would need to run for at least 10.28 hours. I am not a horticulturalist, so I don't know if the DLI is on a per plant basis or a per surface are basis, or if they consider the ppfd of a leaf close to the ground v a leaf near the light source. But i am going to just assume this 2000-3000lumen/ft (14-16 hour) alternative exists because people are lazy and don't wat to do the math, want cheaper upfront costs, and for lack of better documentation.

Given the cost of a greenhouse, its definitely worth asking what exactly is it being used it for, is there an alternate variety that doesn't need a greenhouse, and if it is worth the expense of using it during the off season for a particular function [letting a plant grow, letting it go dormant, not growing anything]. At least in your case, the upfront cost have already been paid for. You probably could get a rough estimate of the cost...err how much space you would have to assign to your compost pile[s] and the cost of compost, but there are still likely to be tons of hidden expenses like irrigation use and equipment, electricity, providing tables/shelves, plant pots and potting soil, fertilizers, pesticides, repairing the covering, etc. Unless it is for personal or scientific/academic value, it's hard to justify the expense unless the marketplace justly rewards it.

Also note that greenhouses, because of the lack of uv rays, tend to reduce the brix number of plants. Typically the higher the brix number the more flavorful the plant, the healthier the plant, and the more insect resistance they have.