4 replies to this topic

[Art Montemayor]

Adam:

“Air-cooled condensers are increasingly common and pose some special control problems. Airflow manipulation can be accomplished through louvers, pitch controls (on the fan blades), or variable-speed motors. Louvers tend to be large, awkward, and easily damaged. The variable-pitch mechanism has been used for control but seems to feature deadband or hysteresis, which can cause limit cycles in pressure and liquid –level loops. The variable-speed motor is the best choice for control, but its expense may not be warranted.

Additionally, most air-cooled condensers consist of several parallel units, covering a large area. Even louvers can be quite costly, and control is usually justified for only one of the units employed. The other fans are switched on or off (as “step-control”), or between low and high speeds, when the signle modulating unit approaches either of the limits of its control range. Shinskey, in “Process Control Systems”; 2nd Ed.; McGraw-Hill”; pp.149-150, describes an automatic system for smoothly coordinating a sequence of on-off operators with a single modulating operator.”

All of the above is a direct quote from Greg Shinskey’s “Distillation Control”, a book that I’ve previously recommended for your reading pleasure. Shinskey goes on to illustrate other solutions and their inherent tradeoffs. But he essentially stays with the expensive VSD as the best operating solution – albeit the tradeoff of money.

I can personally vouch for what he writes. I went through all those learning cycles before he wrote his books based on his Foxboro career experience. I would mention another method that I used that Greg doesn’t mention in his books: an internal recycle process. I employed this method in Cedar Rapids, Iowa on top of the Quaker Oats Mill building – an edifice that was about 8 stories high. I was trying to control the column pressure in a Furfural stripper and if there’s anyone from Iowa reading this, they can appreciate the climate differences between summer and winter in that part of the country. Louvers were a losing proposition. So I built a wooden "box" under and around sections of my Marley air-cooled overheads condenser. The plant worked 24 hrs/day – 7 days/wk, and among the hazards in winter was the freeze-up of the vapor in the air-cooled condenser. I used dampers/louvers inside the wooden structure to recycle some of the warm exhaust air to temper the cold air entering the unit during the winter months and I eliminated any recycling during the summer months. It worked well. This may not be your problem since you haven’t specifically mentioned your ambient temperature characteristics. But it serves to illustrate that you can control the temperature of the condensing process as well as the rate at which you are removing heat. In my time (1970), VSDs were totally out of the question. Today, you have more options.

I think you are on the right track. You have to accurately know the ranges and variations of your local, ambient cooling sink in order to design around it and I think you are attacking the problem the right way. I once again recommend you read Shinskey. If anything, his findings and solutions will fortify your proposals.

Good Luck.
Art Montemayor

[djack77494]

Adam,
There are many installations on the north slope of Alaska where severe weather is obviously a problem. The older air cooler installations tend to utilize adjustable louvers for control. By configuring such a system so that you can adjust the flow of fresh air into, through, and out of the air cooler, you can do a good job of controlling temperature for slow responding systems. I've seen systems where the above is combined with a recycle circuit which lets you cool using a mixture of fresh (cold) air and recycled air. With this, you can ensure that the air is never overly cold and avoid concerns about poor distribution and possible localized freezing. Very flexible, albeit costly.

Louver controls tend to be mechanically complex, somewhat delicate, and subject to hysteresis effects when the mechanisms stick. I think they are not a good choice today, except under special circumstances. Pitch control is also effective though somewhat complex and also subject to hysteresis effects. Both systems can be effectively used everywhere rapid response times are not necessary, in my opinion. Bypass control is simple, fast, and effective, and can be a good choice. My biggest problem with bypassing is that it may be possible to excessively cool (i.e. freeze) product passing through the coolers, especially if multiple parallel paths are available.

I have been thoroughly assured by those who should know that modern VFD-driven systems are cost effective, flexible, and reliable. Obviously, they are also fast responding. As a result, I have convinced myself that this method of controlling air coolers is best for most situations. Where rough control will suffice, consider just using two speed fans.

Regards,
Doug

[ShaunHill]

The points raised so far are all good.
I'm also going to back up to one of the original questions about why bypass control is not good for fin fan exchangers. This is mostly true in situations where you have moisture because bypass control does not protect you from freezing or forming hydrated in the tubes. However, it is an option in systems where you are not concerned about freezing the tubes and I have seen it as a standard method of control for the refrigerant condenser on gas refrigeration systems.

When you are concerned about freezing and are in a cold climate, the best way to go is the type of system Art was talking about. You don't have to build a wooden box anymore (unless you like that sort of thing) since recirculation louvers are a fairly common option. It will cost you a significant amount of money, but it is well worth it if you can avoid a shut-down caused by frozen tubes. All the amine sweetening units that I have seen in Canada include this type of control. The recirculation louvers can be manual (requires operator to go move the louvers when the weather changes) but it works much better with automatic control of the internal air temperature.

In addition to what has already been said about VSD's/ VFD's they have the additional benefit of saving energy and reducing fan noise when the ambient temperature is lower. However, I would still normally include louver control (even though it does have its problems) because fo the limits on turn-down with VFD's.

As to the other question of whether you would put VFD's on all fans or one big VFD to control multiple fans, I would say that I have typically put VFD's on all fans if possible. Although, your actual cost will depend on the standard sizes of VFD's you require. If you want to save money and you have multiple fans, you can put the VFD on one fan for modulating control and base load with the other fan with on/off control. The control for this method is not quite as good as using VFD's for all fans.

[djack77494]

I'm going to challenge the assumptions that VFD drives are expensive and not real flexible. Though I know noone really pushed in that direction, nonetheless those attributes of VFD drives were implied.

I'm told that VFDs are very cost competitive with, if not advantageous to louver or pitch controlled air coolers. Furthermore, I'm told that you can expect accurate and reliable performance of the speed control at least into the single digits (that's based on % of design speed). If that is true, than control should be superlative.

Any comments?
Doug