You have leaks. First try closing isolation valves and see what sections of your system you can pressurize. The start searching for leaks. Try using ultrasonic leak detectors if you are having trouble.
A few dozen small leaks and you would be surprised how much air you loose.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

It won't take a leak of the full capacity of your system to drop the pressure 3 bars.

Bill
--------------------
"Why not the best?"
Jimmy Carter

I understand it won't take a full capacity to drop 3 bar, but why the 4 compressor cannot meet the 30 bar capacity after running for so long.It look like to me that the compressor is only efficient up to 27 bars.

Take the units one at a time and pump into a smaller tank. Record the pump up data to make sure that they are all performing correctly. A bad check valve in one of them could be the cause.
Then once you have PROVEN that they are all working hunt for leaks.
The time does not matter, since the leaks are continuous more time does not help overcome them.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

EdStainless's recommendations are assuredly an excellent starting point to sorting out your problem. I've dealt with similar problems, and isolating and testing the true performance of each compressor was the most valuable step. Pay careful attention to every aspect of the problem. You may find that there may be some conflicts between the control systems of the individual compressors that results in some or all of them operating at reduced capacity some of the time by virtue of their unloading or capacity reducing systems. The details of your piping system may result in slightly conflicting system pressure signals being sensed by the individual compressor's control systems. To get the necessary accuracy to sort out your problem, I recommend using a flow nozzle to measure the isolated flow rather than the more common but less accurate orifice plate meter.

Based on my experiences, it would not be surprising to find that some or all of the individual compressors may be capable of delivering more than their individual nameplate ratings.

Additionally, while you may have leaks in your piping system, your compressed air consuming loads may be drawing more air flow than you are expecting. Metering air flows at carefully selected points in your piping system may prove to be highly beneficial in sorting out your system's true problems.

One of the biggest problems in dealing realistically with plant compressed air systems is that compressed air is very commonly perceived as being "free" when it is actually a relatively expensive commodity.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

I have seen this problem more than once.
The leak is fairly simple.
More difficult is overcoming hubris to get someone to find the leak.
BUT
Measure the current draw of each compressor, if a compressor is not producing rated volume the current will generally be reduced.

Bill
--------------------
"Why not the best?"
Jimmy Carter