There are other more cynical reasons to prefer or require only a single supply to a building.

The electricity supply industry is under pressure to improve reliability of electricity supply. The usual metric is "lost customer minutes" If the supply fails for an hour to six shops, that is 360 lost customer minutes. If instead only a single DNO supply was provided, with sub-metring by the managing agents, then only 60 customer minutes are lost for a similar failure.

There is also pressure to reduce losses in distribution. If a customer can be required to take a supply at high voltage, then that eliminates the transformer losses, at least from the DNO point of view. The losses still occur but are now part of the customers load, and no longer a DNO loss.

If customers with physically expensive premises, such as a large hospital or university can be given one central point of supply, then that removes losses in the perhaps lengthy cable from the DNO account.

Look how well we are doing ! getting greener and more reliable, year on year.

haha - the problem of separating what really should be a joined up view of the losses problem.

As ever thoughtless changes to the rules encourage silly behaviour.

Although  personally I would support the careful use of higher voltage within customer side of large installations (hundreds of meters linear dimensions)  until much nearer the load, as being a very real 'green' solution - perhaps not always at 11 or 33 kV, at least in anything less than the tallest skyscrapers, but certainly encourage more use of 690 and maybe consider permitting 1k2 as 'normal' voltages for large indoor spaces needing a lot of power.  Modern insulation means that while 250V to earth was considered the upper bound in the era of cotton covered rubber and so on, kit for higher voltages is now really not so difficult to manufacture . (and that would mean more use of the U0>400V  column in the table of maximum ADS breaking times for the regs ) electronic sensing of earth leakage and even electronically assisted breakers can make doing  things safely quite practical - the hurdles are more a mixture of designer inertia and regulation.

Mike

haha - the problem of separating what really should be a joined up view of the losses problem.

As ever thoughtless changes to the rules encourage silly behaviour.

Although  personally I would support the careful use of higher voltage within customer side of large installations (hundreds of meters linear dimensions)  until much nearer the load, as being a very real 'green' solution - perhaps not always at 11 or 33 kV, at least in anything less than the tallest skyscrapers, but certainly encourage more use of 690 and maybe consider permitting 1k2 as 'normal' voltages for large indoor spaces needing a lot of power.  Modern insulation means that while 250V to earth was considered the upper bound in the era of cotton covered rubber and so on, kit for higher voltages is now really not so difficult to manufacture . (and that would mean more use of the U0>400V  column in the table of maximum ADS breaking times for the regs ) electronic sensing of earth leakage and even electronically assisted breakers can make doing  things safely quite practical - the hurdles are more a mixture of designer inertia and regulation.

Mike