Extremely sorry for the delay in replying. I was away for a few days.
Is there even a control valve on the discharge side of the pump? From the initial description it sounds as if Danny is simple throwing a switch somewhere to turn the pump on when the roof tank gets low. If that's the case, then wouldn't the check valve be the problem for sure? If the check valve leaks and the system is open at the top, then that's how the air is getting in?
Your quite right. The system is exactly as you say. We have to turn on the pump at the fixed time. We get water from the city's municiple supply for a few hours everyday. We start the pump using device that powers the pump on at a particular time and then powers it off when the tank reaches its overflow level. Due to the need to prime the pump daily, we cannot use this device. Thus requiring manual intervention in an otherwise completely automated setup.
This device would not affect the pipes in anyway. Its completely disconnected from the piping system ( except for the overflow line, which is where a particular attachment is connected which senses water and triggers the device to cut the power to the pump ). Therefore, this device only supplies power to the pump at a fixed time and stops supplying power at overflow.
There is no checkvalve or any form of obstruction in the discharge line from the pump to the roof tank.
If the horizontal suction line is pressurized, then water won't leak back from an unpressured line to a pressurized line.
The horizontal suction line is not pressurised. Water flows in the city's mains using gravitational flow.