Greetings from Seattle on a pretty day.
I just got back from a commissioning kick-off meeting for a new project and thought I would take advantage of some slack in my schedule to get back to my system diagram discussion while enjoying the view from my hotel window.
As you may recall, in my previous post, I had illustrated how you might go about getting started on a system diagram by picking a pipe on the roof of the Pacific Energy Center and simply following it. At this point, we have a page of notes that look like this …
… that covers the portion of the system that is above the roof …
… and is related to the portions of the system diagram that will ultimately be developed as highlighted below.
Note that we have not paid much attention to much of the detail of the system at this point; things like the temperature sensors, make up piping, flow switches, and other devices that you can see on the final version of the drawing, in the pictures I have illustrated the post with, and in the 3D model. Rather we have simply focused on the piping mains.
My reason for developing the diagram in this way (which is generally how I go about doing it most of the time) is that many of the useful things about a system diagram in terms of diagnostics and understanding how a system works are related more to understanding the details of the configuration of the piping or duct mains and not so much to the details of the accessories and trim like drain and vent locations, sensor locations, etc.
Its important to understand that this does not mean those details are not important; they are crucial and may contain the answer to a problem that has plagued the system for years. For instance, a temperature sensor that is on the wrong side of a tee can be the difference between providing the correct information to the control system or the wrong information to the control system, which, in turn, can lead to operation as intended or wasted energy.
But to recognize that nuance, you usually have to have an understanding of the fundamental dynamics of the system and this information is provided by configuration of the pipe or duct mains and the major elements they interconnect (coils, pumps, control valves, etc.) that form the framework of the diagram. Thus, I have found it desirable to develop that framework first and then fill in the details.
That doesn’t mean you can’t pick up the details as you go. There’s nothing wrong with doing that. But for me (this is one of those personal preference things) its faster to get the general arrangement worked out and understand how the system works first.
Having said all of that, I wanted to spend a moment discussing elbows in the context of a system diagram. As the title of the post suggests, in a system diagram,they are generally meaningless and you generally don’t show them. As a rule, I make bends in the pipe on my system diagrams for organizational purposes rather than because there was an actual elbow in the system.
For example, if you compare the piping isometric for the pipe between A and B on the system diagram with the my system diagram, you will discover that the actual pipe has 9-90°elbows and 1-45° elbow between where it comes through the roof ahead of the evaporator and the point where it reaches the first tee after the evaporator. On the system diagram for that same stretch of pipe, I have one 90° turn in my pipe to transition from the vertical to the horizontal to make the “ladder on its side“, an organizational feature.
To say elbows are meaningless may sound a bit crazy, and certainly they are significant from the perspective of the pump and the amount of head or pressure it needs to move water through the system. But they don’t directly affect the path of water through the system in the schematic sense. Thus they tend to be meaningless on a system diagram.
To understand what I am saying, it might be helpful to contrast an elbow with a tee. Both are common pipe fittings and both represent a pressure drop in the system when a fluid flows through them. But, unlike an elbow, a tee represents a point in the system where the fluid’s path can converge or diverge.
If you were water flowing inside the pipe and came to a tee that had one inlet and two outlets, you would have options with regard to where you would go in the system. For water, which is governed by the laws of physics rather than analytical thought, the direction you would take would be related to the physical characteristics of one path vs. the other with the bottom line being that you would follow the path of least resistance.
So for someone trying to understand the system by drawing a system diagram, recognizing that there was a tee in the line and then identifying the characteristics of the piping connected to the two outlets in terms of the valves, pumps, loads, etc. that they contain will be key to understanding the dynamics of the system.
But, in the case of water flowing through a pipe and coming to an elbow, there are no “decisions” to be made, analytical or otherwise. There is only one inlet and one outlet. At the most, it may take more effort (work from the pump) to go the physical distance represented by the length of the elbow’s curving flow path than it would have taken had the flow path been straight. The bottom line from the system analysis and operation perspective associated with a system diagram is that the elbow is simply a bent piece of straight pipe.
Kinda, sorta …
As with most rules, there are exceptions. For instance:
- If the elbow happened to be associated with other elbows that caused the line to drop and then rise again, thereby forming a trap, then, at low velocities, in a system with a lot of sediment carried in the water, debris may build up in the pipe that represented the bottom of the trap, restricting flow more than one would anticipate had the sediment not been there. Eventually, if enough sediment were to accumulate, there could be a mysterious loss of flow in a line that had all of the service valves open and pressure available at the inlet.
- If the elbow was associated with other elbows that caused the pipe to rise and then drop again, thereby forming an inverted trap, air could accumulate and impede flow if the venting was not handled properly. There is an example of just such an issue in a blog post I did several years ago if you are interested in learning more.
- If the system is an open system and the piping rises above the cooling tower basin level, then a lot of the water in the piping above the basin level will drain back down into the basin when the pump stops. And when the pump starts again, it will draw extra water from the basin to fill this piping back up again. Both of these actions can play havoc with the basin level control system if it is not carefully adjusted and if the issue is not considered at design. The result can be basin overflow at shut down and air entrainment and pump cavitation at start-up if the basin level falls too far. In turn, these issues can cause operational problems like excessive makeup, excessive chemical treatment cost, and nuisance problems with chiller flow interlocks when pumps cycle.
- If there were an unusually high number of elbows in one branch of a system as compared to another that was in parallel with it, then the cumulative effect of the elbows might be to cause that branch to have more resistance to flow than might otherwise be expected. This could be particularly misleading if you were analyzing the system dynamics and the branch with the higher resistance due to multiple elbows happened to be the branch that was physically closer to the pumps.
So, since their are exceptions to the rule, I also will occasionally break my rule about not showing elbows and making everything look like a ladder on its side. For instance, I might draw the diagram to show a trap in the line, or put a note on a segment of the diagram indicating that there are more elbows installed in that particular portion of the system than would normally be encountered elsewhere.
The system diagram associated with the air binding problem I mentioned previously is a good example of breaking the rules as illustrated below.
In closing, I want to remind everyone that a lot of the stuff I am discussing has as much to do with personal preference and style as it does engineering and there are not exactly “right” and “wrong” ways to do this. Ultimately, what I try to do is make a diagram that depicts the system in a technically correct manner; the order of connection is right, series and parallel branch arrangements are right, sensors and other control elements are in the correct location relative to other elements in the system like tees, etc.
I also wanted to remind you that the diagrams in this post and other posts in the string are available from the public folder in my Picasa web album if you want higher resolution copies to review.
Senior Engineer – Facility Dynamics Engineering