21 Londonderry Turnpike, Hooksett, NH 03106

Legionella Threat is Real in Hospital “Alternate Water Service Entrances”

Legionella Threat is Real in Hospital “Alternate Water Service Entrances”


In hospitals, an additional culprit for legionellae is the alternate (secondary) water service. The theory is sound: if one of the water services gets shut down (for whatever reason), water will still be available from the second service.

Reality of Alternate Entrances

These alternate entrances usually turn into “dead-legs”; it all comes down to hydraulics:

  1. What is the distance between the two services?
  2. What is the pressure difference at the two services?
  3. What is the actual head-loss through the backflow preventers at each service?  

Because the two service entrances will rarely have the same water pressure, and because the head-loss through the backflow preventers will never be the same, one of the water services will be dominant, leaving the other entrance in the dreaded no-flow or slow-flow situation. 

Example 1: This hospital had two water supplies feeding two separate meters and two 4” backflow preventers. 

What we would like to know:

  1. What is the water temperature of both backflow preventers?
  2. Are the meter readings and flows balanced?
  3. Does the inside of any of the backflow preventers look like a swamp?

Solution: Swap out the large iron backflow preventers for smaller 2″ backflow preventers in parallel (the 3″ copper line is not continuous, and is capped in the middle, out of view).

Example 2: This is a close-up of the gate valve in the headline picture. Yes, it’s entirely clogged! And my apologies if you’re eating.

This hospital had two water supplies:

  • on separate sides of the hospital.
  • a considerable distance between them. 
  • this entrance had a much lower water pressure.

For years, and years, and years, this entrance NEVER flowed water, and consequently became clogged. Everybody had forgotten about it, no preventive maintenance was done; nobody realized the danger to the patients.   Fortunately, the problem was found before the primary entrance was shut-down.

Solutions for Example 1 & 2

As part of an ASHRAE 12 & 188 water management plan, both service entrances need to be addressed. Most hospitals have VFD controlled water booster pump systems, which can be programmed as a quadplex, or controlled through the Building Management System/Building Automation System (BMS/BAS). 

Install duplex (triplex) booster systems at each service entrance, which by themselves, does not solve the problem. However, if the 4 (6) pumps are programmed (or managed through the BMS/BAS) as a single system, then the lead pumps and stand-by pumps can be alternated and sequenced on a 24 hour basis, thus guaranteeing:

  • Water flows through each service entrance
  • Water flows through each pump
  • Back-up pumps are always ready to ramp-up for high demands. 

Example 3: For this hospital, there was a small water demand through the second service, and it was further complicated by dual 4” backflow preventers. 

Solution: The water company thought it was reasonable to install a 2” meter, but no thought was given to the rest of the system, including the legionellae “hotel” backflow preventers. 

Good news: The over-sized backflow preventers were finally changed out for smaller brass 2” backflow preventers. This helped, but did not completely solve the problem. Because the pressures were still unbalanced, only a coordinated booster pump set-up will solve this dilemma.

Day to Day Operations vs. Catastrophic outbreak

If a water management team follows the CDC brochure Developing a Water Management Program to Reduce Legionella Growth & Spread in Buildings, the aforementioned legionella “harbors” will most likely NOT be found:

  • Page 10: does not identify backflow preventer(s), strainers, or expansion tanks in the (1. Receiving) mechanical room.
  • Page 14: recommends to only (a) Check Disinfectant Levels and conduct a (b) Visual Inspection.

Dilemma: All the monitoring will probably come back normal, since the problem is literally hiding in the no-flow / slow-flow “harbors”, underneath everybody’s nose:

  • The water quality is seldom tested in the no-flow zones.
  • The mechanical rooms are usually a healthy 80F+, but the temperature of the backflow preventers, strainers, etc. is never monitored.
  • The INSIDE of the no-flow/slow-flow backflow preventer(s) / strainers etc. is not visually inspected.

So, when an outbreak occurs, it’s always a HUGE surprise, since all of the day-to-day monitoring was normal. And to compound the problem, after a backflow preventer test or jockey pump failure, the legionella “harbor” evidence will most likely get washed away.


  • Further identify and remove legionella “harbors” in the mechanical room (smaller diameter backflow preventers / Gen-5 booster / etc.).
  • Monitor the temperature and visually inspect the inside of equipment to ID these “harbors”.
  • Hospital’s successful water management program must include an aggressive plan for the alternate water entrances