TCO as total cost of ownership was created to compare different cars in terms their costs as fleet cars in the mid 80s. Later Gartner thought it a good idea to move the concept into information technology. The problem with TCO on cars was that you need a very precise number of variables and that a lot of the variables such as the real need of tyres or the real MpG often diverted greatly from the reality, so the car fleet industry has more or less dropped the case for TCO. The only place were you will find TCO as a real concept is the information technology sector, mainly because of the many change in that sector really professional high calibre accountancy is the least established. In controlling it is usually said that any industry employing activity based controlling ABC will not employ semi professional practises such as TCO. A fiend who is a CPA at a medical trust told me once that TCO has the same maturity in accounting as measuring the size of the head to determine the intelligence of a person.
The really funny bit is that if you as an architect will propose TCO as a measure to determine the better solution you will out yourself as only ever have worked in the information technology area. Actually 5 years ago I encountered a business architect that told me off as being too IT centric, to then suggest a TCO 10 minutes later. When I started to smile he could not understand me, until the local controller in a 100 % non IT area just laughed at the suggestion. The point is that if you are asked to do a TCO, do not get too much worked up, as many will see your work work as primitive or semi science. Only if you evaluate products in trials with ABC scenarios in a minimum sampling size you are going in the professional territory, far away from cheap sales techniques.