Delayed Egress Locks

Q: Lately, due to many different construction projects within this hospital, contractors install crash bars (aka panic bars) on doors that have locks for security reasons. The doors do lead to alternate evacuation exits/stairs. The crash bars release the locks in 15 seconds and I have been told that signs notifying people of this is required on the doors. Where is the code for this requirement? One location is an entrance directly into an outpatient care service directly off the public elevator lobby. At two newer locations on another floor, employees are to use their ID badge for access but in one location the sensor is not readily seen. In this location employees frequently open the door via the crash bar setting off the alarm requiring someone to go there to reset the alarm. The message contractors are putting on the doors read: PUSH UNTIL ALARM SOUNDS / DOOR CAN BE OPENED IN 15 SECONDS. The message gives people permission as well as instruction of how to enter a secured area. Where is this code requiring the sign and does it specify the message?

A: Yes… The answer to your question is: Section 7.2.1.6.1 of the 2012 Life Safety Code.

What you have on these doors are called “Delayed Egress Locks”, and the sign that reads “PUSH UNTIL ALARM SOUNDS – DOOR CAN BE OPENED IN 15 SECONDS” is a requirement. If you don’t have these signs, you can be cited by an inspector or surveyor.

Also, as an FYI… you are not allowed to use delayed egress locks on doors in the required path of egress unless the facility is fully smoke detected or fully sprinklered. So, check with your staff to determine if your building is fully sprinklered. In all my 40+  years in this business, I have never seen a fully smoke-detected hospital.

Keep in mind, delayed egress locks are not designed to secure an area. They are designed to allow access through the door on a delayed basis. If the door is located in the required path of egress, then you cannot secure the door, unless it meets one of the exceptions provided in section 19.2.2.2.4.

Magnetic Locks

Q: Is there a code requirement for testing magnetic-locking devices, for a facility maintenance director?

A: There is a requirement in NFPA 72-2010, section 14.4.5 that all interface devices (i.e. relays, control modules) be tested once per year. Since the magnetic locks in access-control and delayed egress locks are connected to the fire alarm system via an interface relay, then the magnetic lock needs to be tested once per year to ensure it disconnects during a fire alarm signal. This test is required to be conducted by someone who is certified in accordance with NFPA 72.

If you are CMS certified or accredited by any of the major accreditation organizations then you would be expected to comply with the manufacturer’s recommendations on preventive maintenance. Most manufacturers of magnetic locks requires periodic maintenance to ensure they are functioning correctly.

Locks on Bathroom Doors

Q: In a multi-tenant office building, can restrooms in the common areas have controlled card access and mag locks tied into the fire system on the entry/exit doors?

A: Well, as long as the locks are installed in accordance with section 7.2.1.6.2 of the 2012 Life Safety Code, I believe it would be okay from an NFPA viewpoint. But you need to ask your state and local authorities to see if they have other restrictions that would prevent this from happening.

Delayed Egress Locks

Q: Our hospital is not fully sprinklered and is not fully smoke detected, but we want to install an infant security locking system in our Mother/Baby unit. I discussed this with our vendor who wants to sell us the infant security locking system, and he says we qualify for delayed egress locks because being 100% fully sprinklered is not the only criterion for compliance. He says we comply because we demonstrate the existence of an approved, supervised automatic fire detection system by having an automatic fire detection system in our hospital, so that should allow the installation of the infant security locking system. The vendor also said as long as the local AHJ approves the installation, that’s all we need, because the local AHJ has the final word. What do you say?

A: NFPA 101 Life Safety Code, 2012, section 7.2.1.6.1 is rather clear: Among other requirements, in order to have delayed egress locks, you need one of the following:

  • The building needs to be fully protected throughout by an automatic sprinkler system, or;
  • The building needs to be fully protected throughout by an automatic fire detection system.

Being fully protected throughout with automatic sprinklers is obvious – you need full sprinkler coverage everywhere in the building. But it appears the term ‘being fully protected throughout by an automatic fire detection system’ is not so obvious. If you are not fully protected with sprinklers, then section 7.2.1.6.1 requires a smoke detector in all occupiable areas. This is explained in section 9.6.2.9 of the 2012 LSC. This means a smoke detector must be inside every room, every sleeping room, every procedure room, every corridor, every office, every conference room, every utility room, every lounge, every classroom, every work-room, every mechanical room, etc. In my 40-years of doing this work, I’ve yet to see a hospital qualify for this in regards to installing smoke detectors in all occupiable areas. If you believe your hospital meets the requirements for being fully protected with smoke detectors, then I would like to schedule a visit and take a look, because I’ve never seen that before.

Please understand the way your vendor described it “demonstrate the existence of an approved, supervised automatic fire detection system”, does not meet the description of being fully protected throughout by an automatic fire detection system. All hospitals have an approved, supervised automatic fire detection system, because the LSC requires that. But no hospital (so far that I have seen) has a smoke detector in all occupiable areas. It’s not required and it is too costly to install. Sprinklers are far cheaper.

Your vendor is correct, though: Sprinklers are not the sole criterion for the installation of delayed egress locks. But, it is one of two criteria, and so far, no hospital is choosing to go with the other choice (smoke detectors). Even if you could afford to install smoke detectors in every occupiable areas, the hospital would likely not be able to afford the maintenance (testing & inspection) and all of the false alarms that go with it.

By the way… the phrase “the local AHJ has the final word” is not accurate. I appreciate the respect that the vendor is trying to say, but all AHJs have the final word, not just the local AHJ. The typical hospital has many (between 5 and 8) AHJs that they have to comply with regarding the Life Safety Code:

  • CMS (Federal)
  • Accreditation organization
  • State licensing agency
  • State agency in charge of hospital construction
  • State fire marshal
  • Local fire authority
  • Local building code authority
  • Insurance company

All AHJs are equal. No one AHJ can override the decision of another AHJ. Any AHJ can decide to interpret the LSC in the way they deem necessary and if it disagrees with another AHJ, then so be it. The hospital must comply with the most restrictive interpretation. So, saying the local AHJ has the final word is not accurate; all AHJs have the final word. For example: If the local AHJ said it is okay to install delayed egress locks for infant security (because nobody wants to see infants stolen), even though the building is not fully sprinklered and not fully smoke detected, that’s not okay with other AHJS like CMS, your accreditation agency and your state agency on hospital construction. So, the hospital cannot do that, because they have to follow the most restrictive interpretation.

I see other hospitals that are not fully sprinklered or fully smoke detected use infant security systems but they do not install the door locking hardware. So, it operates like a warning system. If the hospital does not want to invest in being fully protected with sprinklers (or smoke detectors), then that is their only option. It is an incentive to become fully protected with sprinklers.

Door Lever Hardware

Q: Is there a Life Safety Code requirement for door lever hardware to have a return, so as to not “hook” passing clothing, straps, purses during emergency evacuations? I swear I remember this for healthcare occupancies from somewhere, but can no longer find it in the Life Safety Code.

A: No, the 2012 Life Safety Code does not require a return on door lever handles to prevent hooking clothing during egress. But my good friend Lori Greene (www.idighardware.com) tells me the return is only required by the California Referenced Standards Code, which says: Levers.  The lever of lever-actuated levers or locks shall be curved with a return to within 1/2″ of the face of the door to prevent catching on the clothing of persons during egress. Since this is not a requirement of the NFPA or ICC codes or standards, it would only apply in California.

Access-Control Locks

Q: I have a question regarding access-control locks. I’m told that a PUSH TO EXIT button is required. My question: Is a PULL TO EXIT lever equivalent?

A: No… I would say a “PULL TO EXIT” lever is not equivalent. Section 7.2.1.6.2 of the 2012 Life Safety Code refers to access-control locks which calls for a manual release device, located on the egress side, 40 inches to 48 inches vertically above the floor, and within 60 inches of the secured opening. The manual release device must be readily accessible and clearly identified by a sign that reads “PUSH TO EXIT”. When operated, the manual release device must result in direct interruption of power to the lock – independent of the locking system electronics – and the lock must remain unlocked for not less than 30 seconds.

A “PUSH TO EXIT” manual device is what’s required… not a “PULL TO EXIT” device.

Dead-Bolt Locks on Entrance Doors

Q: A while ago, you made a posting that said a deadbolt lock on a door in the means of egress (such as an aluminum-framed glass sliding door at the entrance of the hospital) would be permitted as long as the door is not a fire-rated door and has no other releasing devices such as a lever, knob or crash-bar. I do not believe that is true, as a deadbolt lock with a thumb-turn should not be permitted in the mean of egress. Please advise…

A: After reconsideration, I believe you are correct. I looked up the commentary under 7.2.1.5.10 in the 2012 LSC Handbook, and it says this about the releasing device on the door:

“Examples of devices that might be arranged to release latches include knobs, levers, and bars. This requirement is permitted to be satisfied by the use of conventional types of hardware, whereby the door is released by turning a lever, knob, or handle or by pushing against a bar, but not by unfamiliar methods of operation, such as a blow to break glass…. The operating devices should be capable of being operated with one hand and should not require tight grasping, tight pinching, or twisting of the wrist to operate.”

It is rather obvious that a standard deadbolt lock with a thumb-turn device does require a tight grasping, twisting of the wrist to operate. So, I agree with you that it would not be permitted in the arrangement as described.

Thanks for bringing this to my attention… I do apologize for the confusion and misinformation.

Stairwell Exit Locked Door

Q: Can a stairwell door that leads to the outside of a hospital be locked with a lock that requires a code to unlock it? I seem to recall that the doors could be on magnets that release upon activation of the fire alarm and that have a touch pad that releases the doors within 15 seconds.

A: No… it can’t. According to 19.2.2.2.4 of the 2012 LSC, doors in the means of egress must not be equipped with a latch or lock that requires the use of a tool or key from the egress side, unless otherwise permitted as follows:

  • Delayed egress locks (7.2.1.6.1)
  • Access-control locks (7.2.1.6.2)
  • Elevator lobby locks (7.2.1.6.3)
  • Clinical needs locks (19.2.2.2.5.1)
  • Specialized protective measure locks (19.2.2.2.5.2)

I don’t know where in the hospital this stairwell exit door is located, but let’s assume it does not qualify for clinical needs locks (psychiatric care patients), specialized protected measure locks (OB, Peds, Nursery, ICU, ER), and elevator lobby locks. That leaves delayed egress locks, which requires the entire building to be sprinklered, and access-control locks which do not lock the door in the path of egress, just in the path of ingress, neither of which allows the use of key-pads to unlock the door in the path of egress.

If you decide to use one of the approved exceptions for door locking, please make sure you read the appropriate section of the Life Safety Code and comply with everything it requires. Most surveyors are pretty well informed on the LSC requirements for door locks and they will hold you accountable.

Strange Observations – Part 28

Continuing in a series of strange things that I have seen while consulting at hospitals…

Dead-bolt locks… Up until the 2012 edition of the Life Safety Code, dead-bolt locks were not permitted on doors in the path of egress in healthcare occupancies.

The reason why is, it would take two actions to operate the door and the LSC does not allow that.

But the new 2012 edition of the LSC changed that a bit and permits existing dead-bolt locks to remain in service as long as the door does not serve more than 3 occupants.

Pharmacy Door Locks

Q: Is it a common practice to keep pharmacy access controlled doors so they will not unlock on a fire alarm activation? Texas Pharmacy board requirements to secure medications is the reasoning behind this setup, which I feel is wrong. Narcotics are secured within the pharmacy area in a narcotics room, so the requirement for securing narcotics and meds is compliant, but I believe the main pharmacy door should release in the event of a fire alarm activation. Our fire alarm system is designed to activate as general alarm so all electrically locked doors other than the pharmacy release. Is there a standard or exception that will allow this type of set-up? The doors can be manually released from the inside, but my concern is the obstacle created by the access control when an emergent response is required.

A: You may be confusing the old NFPA 72 requirement that all doors in the required means of egress equipped with electrical locks have to unlock on a fire alarm signal. At one time, NFPA 72 did require that, but the Life Safety Code always over-rides referenced standards when there is a conflict. In actuality, the Life Safety Code governs in a situation like this. Besides, the technical committee for NFPA 72-2010 made a change to this requirement. In section 21.9.2 of NFPA 72-2010, it says electrically locked doors in a required means of egress must unlock as prescribed by other codes, laws, and standards. This is different than what previous editions of NFPA 72 said, and now NFPA 72-2010 is clear that it differs to other codes, laws, and standards (i.e. the Life Safety Code) regarding the need to unlock a electrically locked door in the means of egress during a fire alarm. There are only five (5) exceptions to 19.2.2.2.4 of the 2012 Life Safety Code, which says doors in the path of egress must not be locked:

  1. Delayed egress locks (7.2.1.6.1) which requires the door to unlock on a fire alarm signal
  2. Access-control locks (7.2.1.6.2) which requires the door to unlock on a fire alarm signal
  3. Elevator lobby locks (7.2.1.6.3) which requires the door to unlock on a fire alarm signal
  4. Clinical needs locks (19.2.2.2.5.1) which does NOT require the door to unlock on a fire alarm signal
  5. Specialized protective measure locks (19.2.2.2.5.2) which does require the door to unlock on a fire alarm signal.

But pharmacies would not qualify for clinical needs locks (used on psychiatric units), elevator lobby locks, or specialized protective measure locks (used on OB/nursery units, ICUs and ERs). That leaves delayed egress locks or access-control locks which must unlock the door on a fire alarm signal. So, your assessment is correct in that the pharmacy door in the required means of egress cannot remain electrically locked during a fire alarm signal. If you examine the Texas Pharmacy board requirements carefully, it probably says the pharmacy must be secured against unauthorized entry… not egress.

What many people forget: The Life Safety Code governs when it comes to the means of egress and over-rides other standards and codes. Getting out of the building in the event of a fire is paramount supersedes other laws, codes or standards. Why don’t you change the locks on the pharmacy doors and eliminate the electrical locks and install standard passage locks (not deadbolt locks) that do not require more than one action to operate the door? This way, the door would not be locked in the path of egress (getting out of the pharmacy) but would remain locked against unauthorized entry during a fire alarm. This is what I see other hospital pharmacies do.