Typical transmission design is a loop for redundancy. Distribution lines will have loops, for the feeders, with an open point such that it is actually a radial

Typical transmission design is a loop for redundancy. Distribution lines will have loops, for the feeders, with an open point such that it is actually a radial system but one that can be fed from 2 substations/circuits. Laterals are the lines where there are no redundant circuits. The other option is networks, which are much more expensive but have much more redundancy. Lateral design: typically a load curve is used to calculate the maximum load for each service point. The system is typically designed to handle close to that amount. If a house has a 200 A main breaker then it can use up to 160 A of current (unless the NEC changed since I did this type of work). Typically a house will never use 160 A even at peak load, but better safe than sorry so a 37.5 kVA transformer is probably used. The power is designed to flow from upstream (substation) to downstream (point of service). The system protection (reclosers, fuses, circuit breakers, etc.) are designed to interrupt over-current from the downstream headed upstream. The exception to this is a network design where protection is designed to protect equipment in both directions. Network design does not happen much at the distribution level, although it is used in major cities. If customer(s) are(is) adding power to the system then at a certain point you will have reverse power flow. And if the system is not designed for this, then bad things could happen and customers are not happy when the electric company causes damages. So the utilities want to be careful. If the rate payers (customers) want a reliable system with the ability to feed power back onto the system then there will be a cost associated with it. For the most part utilities have a monopoly granted by the state and are good at lobbying to protect their interests. What is best for the customer might not be best for the utility.