A REPORT ON
REDUCTION TO
SWIMMING POOL SUCTION HAZARDS


BY
LEIF ZARS

The objective is to avoid I)Suction disembowelment, 2) Suction entrapment, and 3) Underwater hair entanglement.  The procedure I have is regarding the first two - although it could have a significant impact on the third.

It has been stated that a 2.2 pounds per square inch pressure is needed to cause disembowelment.  This is equivalent to a 5.08 foot suction head, (most systems when primed and running will develop up to a 32 foot suction head when blinded off), or 4.48 In. Hg. This proposed system is totally non-mechanical, relies upon the basic laws of physics, and is intended to keep this suction head at or below 4.3 feet.

NSPI codes (as do most other codes) limit suction flow velocities to 6 feet per second in public and semi public pools, and 8 feet per second* in residential pools.  PVC pipe due to its smooth interior surfaces will flow at these velocities with very little actual head.

This limitation gives the maximum allowable flow through the various Schedule 40 PVC pipe sizes as shown in Col. A and B below.

Next let me state that for the below pipe sizes, the friction head loss for 100 feet of pipe when flowing at 6 feet per second is shown in Col. C below, and 8 feet per second in Col. D below.

  A B C D E F G H
Pipe Size GPM GPM* Ft.Head Ft.Head* Ft.Pipe Total Head Total Head* Air Pipe Size
1-1/2" 38 51 8.14 13.86 18.5 2.2 3.8 1-1/2"
2" 63 84 6.08 10.36 25 2.2 3.8 2"
2-1/2" 90 119 4.95 8.42 30 2.2 3.8 2-1/2"
3" 138 184 3.84 6.54 30 1.9 3.5 3
4" 238   2.80   75 2.8   4"
6" 540   1.74   75 2.0   6"
8" 936   1.26   100 2.0   8"


Col. E above shows some typical possible Main Drain pipe footages, including 90s etc. from the main drain sump to pool deck edge.

Col. F & G above shows the resultant pipe head loss due to this footage, plus a 0.7 ft. head loss through the drain opening for 6 feet per second, and 1.23 ft. for 8 feet per second. For example the 3" pipe with 50 feet of length would flow with a head loss of only one half of the 100 foot head loss - or 1.92 Ft. of head loss, again plus the 0.7 ft. head loss due to the drain opening.

What this shows is that with normal Main Drain pipe runs, allowable flow rates of 6 or 8 feet per second can be achieved with a suction head of from 2.0 feet to 3.8 feet (Col. F & G).


Keeping the above 2.0 foot to 3.8 head loss (or foot water heights) in mind, and the previously stated 5 foot maximum suction head, we can see that if we somehow limit the suction head to say 4.3 feet (reasonably below the 5 feet) that we could enjoy the full rated pipe flows in all commercially used pipe sizes and still stay below dangerous suction heads.

So the first realization is that we do not have to place excessive suction on swimming pool piping to effectively gain their rated allowable flow rates.

The next portion of this is of course how to effectively and safely limit this suction head. A Suction Limiting Loop as shown on the attached drawings would accomplish this without the aid of mechanical devices or expensive retrofits.

Going through the following drawings, we can see the development of the Suction Limiting concept using a 42 inch water head.

Drawing "A" shows the "attainable" gravity flow rate into an open container, whereupon it is sucked up by the pump. Attainable is more or less the above figures in Col. A, (or more, as the height of water is increased).

Drawing "B" merely begins to close the open container.

Drawing "C" shows the effect of the larger pipe bringing in more water than the pump is removing.

Drawing "D" moves the pump suction to the side and removes the unnecessary lower tank.

Drawing "E" shows the reduction of the main drain line size and its effect on the water level in the "tank".

Drawing "F" removes the unnecessarily wide upper portion of the "tank", but still leaves it exposed to the atmosphere.

From here it becomes evident that if the pump were to suck more it would draw the "sump" empty thus allowing air to enter the suction line. The basic theory of hydraulics will state that up to this point and once that this happens, the maximum suction possible on the main drain is three and one half feet of water head, none of it caused by pump suction, but rather by gravity alone - just as in diagram A above.

We can completely control this maximum suction head as shown in drawing "J" where we have designed to allow a 5 foot 6 inch suction head (for example only) and drawing "L" where we have limited the suction head to one foot.

The long and the short of the design and installation of a Suction Limiting Loop are as follows:

1. Keep the loop close to where the main drain emerges from the pool so as to reduce the gravity fed portion of this pipe. From the loop on to the pump, it is of course under full available pump suction and can therefore handle current runs.

2. Calculate the pipe distance including the effect of 90s etc. (1 2" 90 is equal to 6' of pipe for example). Then by putting your pipe footage into Col. E above, recalculate Col. F or G, adding 0.7 or 1.23 ft. of head for the drain opening and insure that you do not exceed a 4.3 head. For example, 50 feet of 2" pipe would give a 3.1 foot head loss (50/lOOx6.08), plus the 0.7 ft. head loss from the drain gives a total of 3.8 foot head loss.

3. For retrofit work the typical loops in drawings "G, H, & I" are suggested. They are designed so as to not exceed a 3'6" suction head.

4. Keep your loop simple and straight-forward - note the possible difficulties when redesigning such as in drawings "J", "K" & "M".

5. Regarding the air pipe sizing, experiments have shown the advantage of using the same size vent pipe size as the drain line it is serving. Col. H above will show these.

6. The open end of the gravity loop must be protected from closure and for the moment I have used a glued on PVC cap with 1/8" wide and 1/4" deep saw cuts through it and down the 1/4" cut depth on the side of the cap. This, when embedded in the pool deck or concrete this could perhaps come close to not being easily defeated. I am certain there are other more workable ideas for this.

7. On a typical commercial pool or spa, actual field retrofitting could be inexpensively accomplished by the installation of a 42" long, 1 1/2" vertical PVC loop connected to the main drain line where it emerges from under the deck - burying the loop into a hole at that location.

8. Keep the open end of the air pipe to the elevation of the top of the pool coping or deck.

9. To accurately determine the effective height of a Suction Limiting Loop, a measurement should be taken from the maximum pool water level down to the point where the air pipe allows air to enter the section piping - see H on drawing G - 42".

10. Pump size, line size, plugged drains, missing drain covers - all would be automatically limited to the gravity suction of a 2 to 4.3 foot water head as was built into the gravity loop. Generally it appears that a minimum of a 3.5 foot loop (3.08" Hg) and a maximum of a 4.3 foot loop (3.74" Hg) would work well.

11. In the design of new installations the concept of the Suction Limiting Loop could be easily (and inexpensively) incorporated. Retrofitting is similarly simple but will require digging a hole close to the pool edge.

12. Field tests have demonstrated the validity of this concept.

13. Of course where a pool was constructed with a 1 1/2" main drain attempting to flow at 60 GPM in order to maintain the required turn over rate, this system would be of no help. The pool is definitely outside of the recommended hydraulic parameters, and as such probably should be suitably retrofitted or closed.

14. Hair entanglement could conceivably be lessened by the absolute limitation placed on pipe velocities with this system. Non-the-less, I would favor a snap off grate, one that would come up with the bather's hair if entanglement occurred. And should such a grate remain off for short periods, the danger would be almost nil with the Suction Limiting Loop preventing any build up of suction on the fitting.

HazardDrawA.gif (13791 bytes)

HazardDrawB.gif (14879 bytes)

HazardDrawC.gif (16698 bytes)

HazardDrawD.gif (16013 bytes)

HazardDrawE.gif (15242 bytes)

HazardDrawF.gif (14492 bytes)

HazardDrawG.gif (15549 bytes)

HazardDrawH.gif (12976 bytes)

HazardDrawI.gif (15511 bytes)

HazardDrawJ.gif (17102 bytes)

HazardDrawK.gif (15647 bytes)

HazardDrawL.gif (13326 bytes)

HazardDrawM.gif (15086 bytes)