In order to accomodate their existing system, but try to solve the issue, what I would propose is have wards scale dynamically based on the current consumption. If 85% of all wards are full, add another ward. If the server is running on something like AWS, spinning up a new instance would be quick and easy and could be done real time, and everything else should just be a database update (max ward = X+1). IF they aren't on AWS or a system where such could be spun up on the fly as such, then every time the servers come down (or at least 1/month) perform said check.