Should not be any problem.

I'd like to echo the concerns about adding the safety feedback loop. I remember back in the mid '80's when at a large paper mill a field engineer from a very large controls mfg co. was testing something, I think his logic, and he had switches on the IO modules, I think the actual modules were the yee olde opt22 type with integrated disable switch.

He inadvertently forgot the state of the switch which fires the solenoid to the hatch of a digester which was under test as well. There were painters working underneath that had an untold number of gallons on like 250 deg black liquor (alcohol + caustic soda + wood chips) dumped on them. One died several days later with scaled lungs, the rest severely injured.

Moral of story, if this controls anything that could result in injury, bet to put something like a "Hand-Off-Auto" 3 pos sw so that the auto contact not engaged will inform the SW so that a transition to the auto state and a SW device activate signal will signal some sort of "are you sure?" mechanism before just powering the solenoid.

I agree with kvsubramanyam, adding an external switch should not be a problem.

But are a few cautions: If you add a switch to a separate 24V supply, an unexpected current could flow through the MOSFET's intrinsic diode, nicely shown on your sketch, into the MOSFET's 24V supply. This would happen if that supply was turned off, or if its voltage was more than 0.7V lower than your second new supply. This could cause interesting things to happen to any other items that might be connected to the original supply.

A second caution concerns special driver circuits that use two supply levels, with a higher voltage to start the solenoid and a second lower voltage to power it in a holding state. These come in two types, one that overdrives the solenoid beyond its normal say 8 watts, for a fraction of a second while starting to speedup activation, and a second type that lowers the power level to say 1/2 after turn-on to reduce heating. Solenoid and relay coils operated at their full rated voltage usually run disturbingly hot.

The higher starting voltage is used not only to overcome the dI/dt = V/L inductor-current lag, but also sometimes to create a higher initial current to help overcome the plunger's inertia.

One common example of two-level drive is automotive-engine fuel injectors (waveform from the LM1949 datasheet). The goal is to very-quickly start the flow of fuel into the combustion chamber, while not overheating the solenoid. A secondary goal is to not have the performance too severely degraded under conditions of cold start, where the battery voltage may sag dramatically. Some automotive injector drivers use up to 80 volts for the "peak phase".

These are also an example of another class of solenoid drivers, those that use PWM (pulse-width modulation) to control the solenoid current. Clearly that kind of application wouldn't take well to being bypassed by a mechanical switch. Simply knowing the p/n of the output MOSFET would not be enough to insure that PWM drive wasn't in use, although a simple multimeter check of the turned-on voltage across the solenoid should suffice. In your case if it's a full 24V, then neither PWM or some other type of reduced drive voltage is in use.

It should be fine, but if it were me personally I would add a diode across the bypass switch to handle the reverse voltage from switching the solenoid rather than relying on the IR9530's in built zener diode (especially if the solenoids are large, the reason being the switches lower resistance and contact bounce).

Are there any safety concerns that will be created by bypassing the solenoids in this way? If you don't know, check - the last thing you want is to cause an unexpected operation that could damage the equipment, the user or the product.

Unless the coil uses a freewheeling diode, when disengaging, the mosfet can be stressed by avalanche breakdown just like when it interrupted the current itself, and then some. This specific mosfet is single avalanche rated (500 mJ), but repeated energy should be much lower and it is NOT characterized so it's a bit grey area, and depending on how fast the switch disconnects (relay type, contact materials etc) and the coil energy to be dissipated (1/2Li^2), you might go closer to that than the original design. Mechanical switching is usually slower than semiconductors but that's not always the case. Take mercury switch for example, once used for sub-nano second pulsers. S.M.

I am really sorry that i forgotten to show a freewheeling diode as they are connected across all the coils. So now what will be the situation?

I think no harm. Is it?

If a freewheeling diode is installed, then no problem of the sort I mentioned arises. S.M.

I see there is no problem if you are sure that the coil of the solenoid is good and connect in a parallel with switch a diode in revers bias with high revers voltage typically (RV voltage 200v p-p).

I agree with Redfred I wondered if a double throw switch with center off would be better as shown in my altered copy of the original diagram below. A double pole switch can be used for the extra contacts to operate an indicator for safety as Redfred suggested.

Conclusion.

As there are free wheeling diodes connected across the coils, so there would be no problem to install a separate switch (as shown in picture posted earlier) for switching ON/OFF the coils.

Moreover, during operation through separate switches, i will power off the PCBs having MOSFETs, so will be more secure.

"power off the PCBs" -- Ah, then see my comment #3, where I point out an "unexpected current could flow through the MOSFET's intrinsic diode, nicely shown on your sketch, into the MOSFET's 24V supply". It may be better to just leave the supplies alone.