Should I Use Access Door Position Switches Or Not?

I have the opinion you should almost always use DPS. 

In the case of having limited staff (3 people vs 18k points), having DPS makes triage or establishing criticality of events easier.

For example, if a door is locked, how do I know it is really secured unless it is also shut? (Not propped?)  DPS enables this confirmation.

Even if access monitoring isn't realtime, DPS still confirms doors aren't potentially susceptible to things like tailgating.  

For background, see our post: Door Position Switches (DPS) For Access Control Tutorial

Also, thanks to whoever translated my post title into English from gibberish.

I think DPS should always be used ... it's a critical part of the complete event scenario from a forensic standpoint when trying to figure out what happened at a door, i.e. 1)card scan, 2) Door Open, 3) Rex event (if they are using a motion PIR REX, 4) Door Closed.  Take out #3 if REX device is a button or built into the door hardware.  Same goes for someone leaving .... 1) REX event, 2) Door Opened, 3) Door Closed.   

And more importantly, as Brian said, how will you know your door is actually closed or open? (Let's set aside the nefarious attempt of someone trying to fool a door sensor)

I have a client who keeps some of their access-controlled main lobby doors closed, but unlocked during portions of the day.  Often times, there is no one right inside these doors.  They choose not to use cameras for surveillance for a number of their own reasons, but finally realized they could use the event history of door open and close events to determine if FedEx, UPS, or USPS had actually opened the door and tossed a package inside when they said they did.

If they did have a VMS, they could certainly use one of the readily available systems that offer integration to ACS to link the DPS event with stored video (and not have to worry about setting up motion or line trip recording as the method to determine when someone entered an unlocked door).

Using a DPS provides information after the fact if they choose not to use it proactively “now”.  That may change on a group of doors or buildings in the future and adding them later can be a pain.

Another option if they have alarm systems is to use relay output boards from the alarm system to gather door status into the access control system.

We always specify door position switches as part of access control designs, without exception.

Aside from monitoring the the door for door-forced and door-propped conditions, there can be other reasons for using the DPS. For example, many systems use the  DPS to reset the door unlock time - if you have the unlock time set for seven seconds, and the door is opened sooner, the door will immediately relock when opened rather than staying open the full seven seconds. This can help to reduce tailgating.

Also, if you are using anti-passback, some systems won't register use of the card for entry unless it sees a DPS input immediately after the card is presented.  

This reminds me of being in a Data Center under construction.  Our security is designed for operations, when there are 10-50 around the clock employees who understand the importance of badging .  During construction security gets overwhelmed with 200-300 contractors running around forcing and propping every door they come across while their badge stays in their lunch box.

I agree with everybody, if you don't have DPS's you don't have access control.  Also, if you try to protect everything you protect nothing.  I recommend finding a list of things they do want to protect, and scheduling different doors to alarm at different levels at different times. 

I would get a manageable list, say 20-30 doors, they actually care about.  Server rooms, emergency exits, High Value Storage, ETC.  Set these to alarm at a very high priority, and have alarms of that priority to go straight to the mobile device of whoever is in charge of security or able to respond to alarms.  Make sure that people who do have access to these rooms, hopefully a very small list, are responsible enough to not cause false alarms or can be held accountable if they do. 

With the very high priority alarms out of the way I would then find out if there are certain rooms that aren't expected to have people in them at certain times, but aren't necessarily high priority.  I would set those to a medium priority during slow hours only.  If you get this list small enough and link it up to the cameras, reviewing the alarms the next morning to find out what was going on shouldn't be too labor intensive.

Then you take the rest of the doors, the doors you have pretty much given up on any hope of live monitoring, and set them to log event only.  That way you can run a report, but anybody logging into the ACS won't be flooded with thousands of alarm events they don't care about.  You can also set some of the more burned out doors to free entry during the times they are being abused the most.

"Doors with crashbars and electrified latch retraction kits have a PIR REX burning up the solenoids all over the place." Why is the REX unlocking the door to begin with? A valid read from the card reader should be the only thing firing the solenoid to unlock the door. The purpose of the REX is to shunt the DPS. As always, there are exceptions to this rule but not in this case as I read it...

I am in full agreement of ALWAYS designing and installing the DPS. Without the DPS you access control system is nothing more than an expensive, glorified key. 

Also the purpose of the REX is to log a legal exit event.  The door position switch is part of this logic. For such a low cost part of an access control system I don’t see a scenario when it makes sense from a cost/benefit standpoint