This post looks at why pneumatic controls can be a clue regarding a retrocommissioning opportunity. But before I go into that, I wanted to clarify the difference between pneumatic control and pneumatic actuation as the two are sometimes confused.
Pneumatic Control vs. Pneumatic Actuation
The following slides illustrate a one pipe pneumatic controller serving an actuator …
… and a two pipe pneumatic controller serving a pneumatic actuator.
I loaded the slide deck they came from into my Google Drive so you can access it if you want. If you play the slides as a presentation, they illustrate how these types of controllers work.
The following pictures are close-ups of some actual pneumatic controllers to give you a sense of what this stuff looks like in the real world vs. my simplified cartoon. This first one is of a one pipe pneumatic transmitter, that is in reality, just about as simple as my cartoon. The one on the left is the same as the one on the right but with the cover off, exposing the mechanism.
This picture is of a two pipe controller, which, as you can see, is significantly more complex than its one-pipe counterpart.
This picture is of a pneumatic flow controller (sometimes called a reset controller); they are very common on VAV terminals, including in new facilities where the DDC zone controls have been “value engineered” out of the project to save money (probably not the best move, but that’s a different blog post).
This slide show (new blog technique I just discovered) is of a fairly high end pneumatic proportional plus integral plus derivative controller.
My point in bringing these slides and pictures into this discussion is to allow you to contrast what is inside the controllers vs. what is inside the actuator. The controllers have lots of little tiny moving parts and orifices. These are things that can wear over time and/or become obstructed if the source of air is not maintained in a clean and dry state. Meaning that with out regular attention, pneumatic controllers can drift and loose calibration. Neglected long enough, they will simply stop functioning.
In contrast a pneumatic actuator is just about as simple in reality as it is in the picture; a piston, a diaphragm, and a spring, all inside an enclosure. There is not a lot that can go wrong and if something does go wrong, someone with basic mechanical skills can probably fix it.
Pneumatic actuators, when contrasted with electrical actuators, can offer a lot of advantages including a lot of actuating power for the dollar, simplicity and reliability, and speed. All of these things can be an advantage in the context of a control system design. As a result, I actually prefer control systems with pneumatic actuators and DDC front ends, at least for the central plant and air handling equipment, where speed and power are highly desirable. At the zone level, using electric actuators solves the pneumatic tube leakage problem and is my preferred choice.
For those who want to know a bit more, I wrote an article for Networked Controls a while back that contrasts current technology actuators and sensors with the pneumatic technology and you can pull it back from my Google Drive if you want to read it. If you want to know a bit more about pneumatic controls or controls in general, then there are a number of resources in the resource list I post on the blog occasionally with the the Honeywell Gray Manual being one of the best.
The General Case
There are a number of things about pneumatic controls in the general case that make the systems they serve likely retrocommissioning targets. One is their mechanical nature, as I discussed in the previous section.
Another is that they are not very well understood and it takes a certain level of expertise to be able to work with them. That expertise is not particularly difficult to pick up, but there are fewer and fewer people with direct exposure and experience these days. So if you discover a facility that uses them still, then the chances are higher than they used to be that the technology is not understood very well by the people charged with operating and maintaining it.
That leads to another reason that pneumatically controlled systems become good retrocommissioning targets; if something is not understood, it is likely that it is also not maintained as well as it could be. Combine that with the mechanical nature of pneumatic controls and you have some significant potential for things to be out of adjustment.
Compounding the problem is that it is becoming harder and harder to find replacement products. And, many of the better quality ones are no longer manufactured due to lack of demand. For instance, I found Powers RL243SW pneumatic switching relays to be reliable and well constructed. But the last time I tried to buy one, I was told that they were a discontinued item, meaning I had to use something of lesser quality to replace that function. (Incidentally, the product bulletin that the link takes you to has a cross section of the relay, which will give you more insight into the mechanical nature of this type of equipment).
Our Target Facility’s Case
As I mentioned in the preceding post, I found the control drawings for the original project at the back of the original contract documents and discovered that the control system was originally laid out as a pneumatic system.
There are some pretty good retrocommissioning targeting clues regarding the design intent in these drawings too, just like the others we looked at. But in this case, they are clues about the potential for dysfunctional operation in addition to some clues about how we could save energy. Here is the list I came up with as I looked at the design along with an explanation of why it is a clue.
The System is a 100% Outdoor Air System with the Freezestat Located Ahead of All of the Heat Transfer Elements and Set to Trip and Shut Down the System at 36°F.
Think about it; this system, by design will handle sub-freezing air. That means that if the controls were installed as described on the contract documents, you would not be able to keep the unit running any time it got below 36°F. Clearly, that will not work.
Its hard to say what happened until you get out on the site to look, but this is a clear indication of a problem that required some sort of field revision. That, in turn, is a clear indication of a potential performance issue that could have been solved in an energy intensive manner. Thus it is an indication of a potential RCx opportunity to solve a problem, improve performance, and save energy.
There is a Mixed Air Control System Shown in the Control Diagram that is Not Mentioned in the Control Sequence
There are a couple of issues with this particular observation. One is the obvious one; the diagram doesn’t agree with the narrative.
Another is that for a 100% outdoor air system, there really is no mixed air. So, if that particular controller were installed, and set to maintain the cooling coil discharge temperature (52°F; economizers are a cooling process and are typically coordinated with the mechanical cooling to maintain about the same discharge temperature), then as the outdoor temperature dropped below that value, the controller would likely try to close the outdoor air damper to the minimum position, which would restrict air flow into the system.
The Outdoor Air Damper is Closed When the Unit Shuts Down via an EP relay that Shuts Off Main Air to the Controls and Bleeds Pressure from the Control Panel
You have to look closely at the drawing to see this and know a few conventions about pneumatic controls. For one thing, you need to know that EP stands for Electro-Pneumatic relay. Basically it is an electrically switched pneumatic relay that typically has three ports (the link takes you to my favorite one).
In applications like this one, it is typically used to apply or remove pressure from a device or collection of devices. In general terms, it is quite desirable to do this to the outputs from a pneumatic panel to return them to their normal state when a system is off. Typically, the actuators are configured so that their “normal state” places things in a safe position.
In our case, the idea would be to remove pressure from the outdoor air damper to cause it to close when the fan was off. This shows up in the narrative sequence under Freeze Protection.
It might also be desirable to set up the system so the the hot water valve fails to full heat to help protect the coil. You have to be a bit careful with this though because in the off state, all of the air in the fan unit will be heated to the temperature of the supply water. So, when the fan starts, you blow a slug of hot air down the duct. If the air is above 165°F and there is enough of it, you might melt the fusible links in your fire dampers, causing a bit of an operating problem until you can get them reset.
Here’s a picture of a couple different fusible links in case you have not seen one before, including one that melted one day …
… specifically one Christmas Eve when I was a facilities engineer at KSA (see Extreme Weather; Your Winter Wonderland can be a Facilities Engineer’s Worst Nightmare for a broader perspective) . It was after Home Depot had closed for the holiday, so we couldn’t run out and pick one up in the fusible link aisle; kind of a problem. But I digress.
If you look at the pneumatic control drawing, you will notice a device labeled Fan EP Relay. On the left side of the device, there is a circle with an “M” in it which stands for “Main Air”. In pneumatic jargon, “Main Air” means steady, stable, clean, dry source of air pressure, probably in the 25-30 psig range. 30 psig is usually the maximum allowable pressure for commercial pneumatic control devices. Typically they can operate with 20 psig air and supplying it at 25 psig or higher provides some margin for error and pressure drop in the distribution system.
If you look to the left of the EP, there is also an “M” but with a “1” subscript. This implies that a circle with an “M” and a “1” is an air main that is switched by the fan EP.
Now, notice that source of air pressure for the two controllers associated with the fan is the switched main air. This has a subtle but significant importance. It means that the pressure to the outdoor air damper and control valves is being bled off any time the fan shuts down. The good news in that is it means that the outdoor air damper is being closed when the fan is off (assuming it is arranged so that the motion that occurs with no air pressure to the actuator closes the damper; the narrative requires it but the drawing is not clear on this and it would be better if it was). Similar things happen with the coil control valves, but we are left in the dark as to the specifics and they are not addressed in the narrative.
The bad news is that the pressure is relieved from the actuator by relieving it from the controllers. Specifically, instead of providing an EP switch for each controller output, which would have required two EP switches (read more cost if you are a control contractor bidding the job and have to get low to win) the drawing (and successful low bidder) would only provide one EP switch.
The reason this is bad news is it means that any time the fan cycled, air pressure is bled from all of the controllers in addition to the actuators. Not a problem for the actuators because they are very simple devices and are designed to work with significant variations in supply pressure. In contrast, the controllers are complex devices with a lot of springs and levers and diaphragms that would be better off if you kept a fairly steady pressure on them.
If you cycle the pressure, the effects of hysteresis can set in. Meaning that when pressure is relieved and then reapplied the various diaphragms, springs, and levers come back to a slightly different position from the one they were in before the big pressure change. Meaning that the controller is no longer calibrated. Some controllers tolerate this better than others. But I have worked with controllers that would be useless in terms of calibration after one cycle.
So, the bottom line for me on this one is that if I see a system where one EP can remove pressure from the entire control panel every time something like a fan is shut down, then it’s a red flag regarding the calibration (or potential lack there-of) for the controllers in the panel.
The Control Diagram and Sequence Show that the Two Coils in the MAU are to be Controlled in Sequence Based on a Space Condition
The issue here is that the arrangement puts a pretty good time lag between when something might change at the unit and when that change would actually be detected and reacted to by the control system.
Consider what would happen if you tried to start up this system on a sub-freezing day; lets say it’s 30°F outside. And we will assume that the unit has only been off for a short time; maybe the operators turned it off to replace the filters. So, the space where the space thermostat is located is under control. Meaning its not cold or hot.
So, when the operators turn the fan back on after the filter change (assuming the outdoor air damper issue alluded to above had been addressed and the freezestat had been located to be downstream of the preheat coil so that it doesn’t trip as soon as it sees cold air), the outdoor air damper would open and 30°F air would immediately be introduced to the unit.
But, even though it is below freezing right away at the MAU (and associated water filled coils) the space temperature is just fine. So, the controller in the space may not even be calling for any heat. In fact, if there were enough internal gains, the heat that built up while the unit was off could mean that it is calling for cooling.
And, to have a meaningful impact on the space temperature, the cold air coming in at the unit will have too cool down the mass of the system, including the water in the coils. Its not out of the question that it will cool the water in the first couple rows of the preheat coil and maybe the cooling coil down below freezing before it cools the space to the point where the controller asks for some heat.
All of that means that the preheat coil could see freezing air for a significant amount of time with out much if any hot water allowed into the circuit until the space starts to cool down. If you are lucky and the temperature in the discharge of the preheat coil (where the freezestat should be) drops below the setting of the freezestat, the unit will trip off. Annoyed, you will reset it and the cycle will repeat until the cold air gets to the space and cools it enough to get the hot water valve open, allowing the preheat coil to work and keep the freezestat from tripping. Hopefully, all of that happens before you freeze a coil.
If you are unlucky and some other operator who didn’t understand exactly what was going on got tired of all of that monkey business (usually a more colorful description will be applied to the problem) and wired the freezestat out of the circuit, then you will probably freeze the coil, something that will not become apparent until after the coil thaws out, at which time, you will have, as Jay Santo’s would say “a significant emotional event”.
The Preheat Coil May be Installed Up Against the Cooling Coil with No Space for a Freezestat (or Preheat Coil Controller)
The control schematic shows the preheat and cooling coils butted up against each other. The actual drawings are not quite as clear. But the actual drawings also do not clearly ask for them not to be butted up against each other. So, if the job was not specified that way and then was put out to bid with the award going to the low bidder, it is unlikely that the units were provided with an access section between the coils.
This makes it harder to clean and maintain the coils. And it makes it very difficult to install a sensor between the coils. Sadly, for a preheat coil (technically, a coil that by design, must handle sub-freezing air) to work, both the freezestat and a sensor for a controller dedicated to the preheat coil need to go there.
The preheat coil’s basic function is to protect the rest of the system and facility from sub-freezing air. From a control standpoint, to do that effectively (in addition to being physically configured to do it) the coil has to be able to react quickly to changes in outdoor temperature before they can freeze the coil itself. That means that the coil needs a dedicated controller that maintains the discharge temperature from the coil (all though I did see a feed forward strategy applied one time to solve the “no space between the coils” problem).
From a system protection standpoint, the freezestat needs to be after the preheat coil because by design, the preheat coil will see subfreezing air. But it also needs to be ahead of everything else since by design, everything else is not designed to deal with subfreezing air effectively.
The Control Sequence Operates the Circulating Pumps Continuously
This is highly desirable when the weather is below freezing but generally not necessary when a coil is not active (i.e. when the outdoor temperature is above the preheat coil discharge temperature). And it may not be necessary at all for either coil if the temperature is above freezing and there are not laminar flow issues at low load.
While the pumps are small, shutting them down when they are not required will add up; basically a little bit over a long time can add up to a surprisingly big total. For an example of this, take a look at the post titled A Little Bit Over a Long Time Equals A Whole Lot.
The Pneumatic Control System Bottom Line
The bottom line is that for our target facility, the possibility that the MAU is managed by a pneumatic control system raises a number of significant performance and energy conservation issues, enhancing the MAUs status as an RCx target. And since there are a number of similar systems, what we find by looking at one in depth will likely ripple out as improvements in all of the systems, compounding the savings and the benefits of our efforts.
Senior Engineer – Facility Dynamics Engineering