Actual flow rates? HELP!

By Timothy Allinson, P.E.,
Murray Co., Long Beach, Calif.

It’s no secret to many of us in this industry that the systems we design are often dramatically oversized. Sometimes this is the fault of the designer, who adds factors of safety in fear that things might not work. Even those with the courage and faith to design to the knife’s edge of the Code end up with systems that are oversized, because the Codes are theoretical rather than empirical, and what little empirical data that might have served as their foundation at one time, such as Hunter’s Curve, is ancient by today’s standards.

This fact recently came to light when talking to Bob McCook, an industry friend and sales rep, about water softening systems and flow rates. He mentioned in passing that he had sold the water softening system to the Bellagio Casino in Las Vegas. The system measures and records flow data and serves the entire 900-room casino tower with cold and hot water, except for the cooling tower, and recorded the peak domestic flow rate as 190 gpm.

Think about that. A quick, low-end estimate based on one toilet, two lavs and a regular bath/shower in each room would yield a flow rate of nearly 1,000 gpm; that’s not even including the penthouses, with more fixtures and high-flow tubs. In reality, the water demand was, reportedly, only 190 gpm. Something is seriously wrong with this picture.

The likely explanation is that, in high-end casinos, guests spend very little time in their rooms and shower at odd times, creating a very low diversity. If you’ve ever been to Vegas, you know what I’m talking about. But the disparity between the Hunter’s Curve flow rate and the measured demand warrants serious consideration.

I know that there are people in our industry working on updating Hunter’s Curve and fixture unit assignments for ultra low flow fixtures but, given this extreme disparity, these seem to be baby steps. As engineers, we rarely have access to after-the-fact empirical data of actual flow rates for the projects we design. This is what we need more of — empirical data. Some of you out there have it, and those who do probably have a limited amount of it. There are probably manufacturers, like my associate Bob, mentioned above, who have more of it; this is what we need.

Calling all manufacturers! We need your empirical data!

Here’s an example of some valuable empirical data I recently came by. Most of us know rules of thumb for estimating cooling tower make-up, a common one being 3% of the tonnage. For example, a 1,000 ton tower would require a make-up flow rate of: 0.03 x 1,000 = 30 gpm

In this example, a make-up line sized for 30 gpm would satisfy the tower at peak demand, and the basin would not run dry, even on the hottest of days. But how much water will that tower use in a day? Would it be: 30 gpm x 60 min/hr x 24 hr/day = 43,200 gal?

That seems like an awful lot of water, doesn’t it? If you had to size a water tank for a three-day supply, as we often do here in So Cal for hospitals, would you assume 130,000 gallons just for the tower?
To answer this question, an HVAC friend contacted a local area hospital where he knew the facilities guys quite well; their towers, of course, had a water meter. Based on the data we were able to obtain, during a very hot week (temps in the 90s and dry), the hospital cooling towers consumed the calculated peak for nine hours a day. In our example, the actual daily water demand of the 1,000 ton tower would be: 30 gpm x 60 min/hr x 9 hr/day = 16,200 gal

The three-day supply required for the hospital’s tank would be 48,600 gallons. This is a far cry from the calculated 130,000 gallons. This, of course, makes sense. The towers would not operate at their peak all night long. The calculated peak make-up flow does not run continuously but will cycle on and off all day and night, more during the day and less at night, the sum total equating to nine hours of peak flow. That was a brilliant piece of empirical data to stumble upon, and it made me thirst for more of it.

Back in my NYC engineering days, I designed a replacement pumping plant for Bankers Trust Plaza, which later became famous because of its controversial demolition as a result of structural compromise from the 9/11 bombings. The replacement pumping plant had a flow meter, but it was not a recording meter, and there ensued some debate over the ideal flow and pressure settings for the lead and lag pumps. As a result, I had one of my engineers sit by the pumps for an entire day, manually recording the flow and pressure every five minutes, as well as the pump sequencing.

The domestic water pumps at the plaza also fed the cooling towers on the roof, which, for this data-heavy building, were large, although I don’t remember their tonnage, and they are, sadly, long gone. Because it was an office building, the water demand was not as intensive as that of a casino or a hospital, but it was a cubic tower with a city block footprint and 40+ stories, so there were a lot of fixtures.
At the end of the day, the water demand was seen to flow at a fairly steady rate of 150 gpm; this flow rate represented mostly the cooling tower demand. The flow for this building was what I would call very spikey, with frequent spikes in flow of 300 or 400 gpm that were too brief to kick on one of the lag pumps. Once a lag pump did kick on, it would run for a considerable length of time as the brief spikes would prevent the minimum run timer from timing out.

This building had originally been provided with three 100 HP domestic water pumps. As part of the pump replacement, we added a 40 HP lead pump. At the time, variable frequency drives were still impractically expensive for domestic water applications. Today, I would have stayed with the three 100 HP pumps and added VFD drives. This would have smoothed out the pump controls a great deal.

Details of the pump replacement aside, this project served as an example of a building that consumed far less than its calculated demand. The combined total of 340 HP rarely used more than 40 HP and when it did, it never demanded more than 100 HP. I have since heard many stories of pumping plants that never use more horsepower than the lead pump.

Timothy Allinson is a senior professional engineer with Murray Co., Mechanical Contractors, in Long Beach, Calif. He holds a bsme from Tufts University and an mba from New York University. He is a professional engineer licensed in both mechanical and fire protection engineering in various states, and is a leed accredited professional. Allinson is a past-president of aspe, both the New York and Orange County Chapters. He can be reached at laguna_tim@yahoo.com.