In a post earlier this week, I showed you the clues that led the commissioning team I am working with to wonder if we might not experience a few condenser
water system operational problems as we brought a new central plant on line.
Our curiosity was initially triggered observing the piping geometry on the roof combined with our knowledge of what was going on below the roof. The small scale illustrations below, which were included at full scale in the original post, will refresh your memory.
As you can see, the pipes rise to above roof level, then drop back down through the roof to go under a roof level walk-way and then re-emerge under the cooling tower basins. This configuration forms an inverted trap. Such a configuration is often a good place for air to collect, with little natural inclination to move on. Once “trapped” in the line, it can impede flow, triggering all sorts of operational problems.
As you can see from the photo and diagram, the contractor installed manual air vents to facilitate the removal of air from this high point when the system is filled. But a combination of factors – several of which were presented as clues in the previous post – caused us to be concerned that in this particular instance, air removal may not be as simple as that.
Granted that the condenser water system is an open system and thus less likely to need automatic air venting. But, in this particular system, the velocities will be extremely low through the inverted trap if only the modular chiller is running, and to some extent, it is velocity that helps to carry air through a system to a point where it can be vented, either naturally or manually.
Now, lets put the clues together to understand the specifics of the potential problem that the commissioning team was concerned about.
- Air is soluble in water, as evidenced by the picture of the glass of water; when you draw a glass of cold water from the tap and then let it sit around and warm up, water bubbles form on the sides of the glass.
- By its nature, the water in the condenser water system
will be exposed to a lot of air due to dissolved air in the makeup water and due to the fact that it is spread out and splashed through the cooling tower fill while convection and the fan move air past it.
- When the water in the system is warmed up, some of the
air will be driven out of solution, either because the system was filled with relatively cold “city water”, or the system has been idle and then placed on line, or because of temperature changes in the system.
- Pressure changes that occur as the water moves through the system will have a similar effect, especially in the pump impellers where velocities tend to be high and thus, the pressures low. Here is an illustration of pressure vs. velocity through the eye of an impeller from the Pump Engineering Manual published by the Durion Company that illustrates what I mean.
- Air tends to accumulate at the high points in systems, which, for the system we are discussing, is the inverted trap at roof level and the cooling tower basin. The top of the volute on a horizontal split case style pump can also represent an intermediate high point in most systems, which is why they are often provided with a tapped opening which can be fitted with a manual vent. The pumps on this project are just so equipped, as illustrated below.
- At full load, the velocities in the system are probably high enough to carry the air through the trap and into the cooling tower on the return side of the system (from the load to the tower). But if only the modular chiller is in operation, the velocities are extremely low in the large 14″ lines and the commissioning team was concerned that significant air could accumulate in the inverted traps. Once it did, the low velocities may mean there there is not enough energy available to carry it out of the system. As a result, the air pocket would begin to obstruct flow.
- On the supply side of the system (from the towers to the load), things could be more difficult because for the air to be carried out of the system, it needs to be forced all the way down through the building and back out again. This would allow it to accumulate in intermediate high points like the tops of pump volutes and also would compound the problem on the return side.
As a result of the teams concerns, an RFI was generated (a formal Request For Information used to document questions and answers on a construction project) asking if it would be desirable to install an automatic air vent on the inverted trap in place of the manual air vents, that were provided. The designer’s observation was that air venting is usually not an issue in an open system and thus, they felt the auto air vents may not be required.
Since the installation complied with the contract document requirements, adding the automatic air vents would have represented a modest change order. So, given that we were close to start-up and, as Jay Santos is fond of saying “nature doesn’t lie”, we decided to simply wait and see what happened. No sense adding a part that may not be required. And, since we had already thought through what the issues might be, the commissioning team had an idea what to look for and where to look for it if problems did occur.
About a month and a half after I took the picture at the beginning of the post, the systems were filled and started up. During the course of the start-up, nature spoke to us. Specifically, the contractors were in having problems with repeated air accumulation in the system. Despite diligent efforts to follow good practice for flushing, filling, and venting the system, they continued to encounter problems with air being trapped in the top of pump volutes and in the inverted trap, especially if they were only operating the pumps for the small modular chiller.
After revisiting our concern, as discussed previously, everyone agreed that the commissioning team’s suspicions had been confirmed and the paperwork was initiated to add two automatic air vents, one on each inverted trap. Stay tuned and I’ll let you know how well they work in a future post.
Senior Engineer – Facility Dynamics Engineering