Dissolved Oxygen control at WWTP

P

Thread Starter

Patrick Keough

Does anyone have experience with Controlling Dissolved Oxygen for wastewater. We are trying to generate ladder logic to modulate blower valve positions, which will control the D.O Level in the aeration tanks. We have used PID control before but that doesnt seem to work to well. We thought of using a timer to control the signal to the valve. We've done this to allow the process to"catch up" with the changes in valve position.

Does anyone have any ideas or solutions??????
 
--------
This is a non conventional control loop.
We made similar job using:
1_ Orion analyzer
2_ Digital multiport valves
3_ Micon controller
 
J

Jake Brodsky

What kind of criteria are you trying to meet?

Do you know that you have enough capacity in your blowers and difusers to properly airate the solution?

Why didn't the PID loop work? What did it do?
 
Patrick,

I've done a couple of these in the past. I'm not sure what problems your having, so I'll just describe the kind of system that I've been involved with.

The Aeration Blowers are used to maintain a constant header pressure. The kind of controls used depends on what kind of equipment implemented.

Modulating control valves (butterfly) and air flowmeters (thermal dispersion) are installed between the low-pressure air header and each of the aeration basins. Using cascade-type loop control, the DO signal (outer-loop), writes an air flow setpoint (inner-loop) that is used to control the position of the modulating valve.

Because of the type of process, I've found that the DO control loops are pretty tough to tune.
 
P

Piotr Kowalski

On my wastewater plant I have 2 blowers for each SBR and control is
done by simple switching one or two of them on and off, depending on
D.O. level in the reactors. Because of the big inertia of the system and
proper hysteresis we don't encounter frequent switching of the blowers and
moreover if they run, they run at 100% capacity (good for saving energy).
Oxy level is not as stable as with continuous control, of course but is
good enough for the process in our SBR's.

Regards
Piotr Kowalski
 
A

Al Pawlowski

I missed the original post so please excuse this post if it has nothing to
do with the original. However, I have some experience with DO control in WW
aeration and hope it is useful to someone.

1) The DO control itself is usually pretty straight forward. Use a PI
controller to adjust the air flow to the area where the DO is being
measured. If you have the typical multi-pass tank with separate butterfly
valves to each tank, it is appropriate to use a DO controller cascaded to
separate air flow controllers on each tank; helps tuning to decouple the
effects of different loadings. If you have a DO controller for each
tank/pass, the cascade is not necessary. However, it is also common to
average, or low select, signals from multiple DO analyzers. My experience
is that it is pretty easy to get good DO control (say +/- 0.1ppm) for any
area with an analyzer and feed valve. Because of cost/benefit, however, the
most I have seen in one pass of a tank is two controlled zones. Obviously,
if you had many zones in one pass, there could be problems with controller
interaction.

2) It is typical to include energy management control on the blowers
providing the air. Typically, there are multiple units and some adjustment
on how much air the individual units supply; for example, VFD's on motors
or adjustable suction valves. Usually, the units discharge into a common
manifold which is then split to separate tanks. The most common control
scheme I have seen tried is to hold a set discharge manifold pressure (just
high enough to supply the needed air). I have never seen one of these
pressure schemes work. There are too many other variables (besides air
demand) that affect the pressure. What I have been successful using is to
monitor the air flow control valve positions and provide just enough air
(blower output) to keep the most open valve mostly open, but closed enough
to get some control; say 60% +/- 5%. If the most open valve goes too open,
boost the blower output (more units, more suction, more speed, whatever
control you have) and vice versa. Often the blower control also includes
alternation of the on-line units similar to pump station control.


Al Pawlowski, PE
[email protected]
dba ALMONT Engineering
Baton Rouge, LA USA
 
O

O'Connor, Denis

Does anyone have recommendations on self-cleanding DO sensors?

Denis O'Connor
Manager of Operations
Alliant Energy Integrated Services - Cogenex
Boott Mills South, 100 Foot of John Street
Lowell, MA 01852

tel 978.441.0090 ext 240
fax 978.441.4626
E-mail: [email protected]
 
P

Piotr Kowalski

Hello!
Al wrote: "The most common control
> scheme I have seen tried is to hold a set discharge manifold
> pressure (just
> high enough to supply the needed air)."

Sounds strange to me. Pressure in the manifold is constant and equal to the water level in tank plus preasure loss on diffusors. Forcing higher pressure will result in higher air flow and possible damage of diffussors. Moreover efficiency of the process depends on water level (the longer way for bubbles the more oxygen from air is dissolved in water) and water temperature (increasing temperature prohibites oxygen dissolving process) so trying to control pressure seems to me to be too indirect way for controling D.O. concentration. Notes above applies to SBR system with elastic diffusors and roots blowers.

Best Regards
Piotr Kowalski
 
P

Patrick Keough

Jake, what we have is a typical multi-pass tank with separate butterfly valves to each tank,(3 valves, 1 to each tank)
Each tank has its own D.O sensors, and what we want to do is to control the Dissolved oxygen level at a specific setpoint(lets say 1.5ppm with a 0.2ppm deadband). There are 3 existing blowers that run continuously, we need to adjust the air flow to each tank by adjusting the valves, which in turn will control the D.O level in the tank. We have found in the past that because of the process it is extremely difficult to tune the PID loop.

In short our PROCESS VARIABLE is the D.O in the tanks, SETPOINT=ppm, and the CONTROL VARIABLE adjusts the valves to keep the DO within range.

Hope that gives you a better understanding. Again any help is appreciated!
 
K

Kiel-1, Larry

Hello. Several years ago I was researching Dissolved Oxygen sensors for the
2 small treatment plants we have here on site and I found a self cleaning
sensor distributed by Great Lakes Instruments (GLI). The sensor is actually
called a Zullig D.O. sensor and it uses a small motor with a rotating
diamond grind stone to clean the electrode surfaces that it uses to detect
the D.O. level. When I was doing my research I talked to a couple of plant
supervisors in other states who had the sensor installed at their sites and
they were very complimentary about the accuracy, lack of calibration needed
and overall performance of the unit. I specified the combination of the
sensor and analyzer for our plants but the project was never implemented so
I don't have any actual experience with the unit.

GLI also makes a standard membrane D.O. sensor (model # 5440D) that uses an
air blast to try and clean the membrane. I didn't do any research on that
model so I can't advise on how well it does or doesn't work.

You can reach GLI @ 414-355-3601. I'm afraid I don't know their web site.

Larry Kiel
Kennedy Space Center
321-867-1415
 
K

Kiel-1, Larry

Hello. Several years ago I was researching Dissolved Oxygen sensors for the
2 small treatment plants we have here on site and I found a self cleaning
sensor distributed by Great Lakes Instruments (GLI). The sensor is actually
called a Zullig D.O. sensor and it uses a small motor with a rotating
diamond grind stone to clean the electrode surfaces that it uses to detect
the D.O. level. When I was doing my research I talked to a couple of plant
supervisors in other states who had the sensor installed at their sites and
they were very complimentary about the accuracy, lack of calibration needed
and overall performance of the unit. I specified the combination of the
sensor and analyzer for our plants but the project was never implemented so
I don't have any actual experience with the unit.

GLI also makes a standard membrane D.O. sensor (model # 5440D) that uses an
air blast to try and clean the membrane. I didn't do any research on that
model so I can't advise on how well it does or doesn't work.

You can reach GLI @ 414-355-3601. I'm afraid I don't know their web site.

Larry Kiel
Kennedy Space Center
321-867-1415
 
A

Al Pawlowski

I would not say it could not be done (it could, but to do so would indeed
be "strange"), but the pressure control I described in my previous post was
not for DO control. It was to minimize blower energy. The theory is that
the pressure controller would adjust blower output to keep feed pressure
at the maximum ever needed by DO controller/s. Unforunately, that maximum
varies a bit and the blower energy required to hold it is very sensitive to
pressure.

I mentioned it because I have seen it implemented, and be abandoned, in
many plants. I have never seen it better the average plant operator.

Also, while the piping, and diffuser, loss due to flow may be small
compared to the water head, it may still be significant. I have seen more
than one case where it was enough to severely limit the expected blower
output of a system.


Al wrote: "The most common control scheme I have seen tried is to hold a
set discharge manifold pressure (just high enough to supply the needed
air)."....................

Piotr Kowalski replied
Sounds strange to me. Pressure in the manifold is constant and equal
to the water level in tank plus preasure loss on
diffusors......................



Al Pawlowski, PE
[email protected]
dba ALMONT Engineering
Baton Rouge, LA USA
 
D
Continuous control of DO is difficult due to the long dead time or the time that a change in the output takes to effect the process variable. Consider using a loop tuning software package.

 
Top