Desktop version

Home arrow Environment

  • Increase font
  • Decrease font


<<   CONTENTS   >>

Liquid Circuit Problems

Pumps are the most common liquid "movers" in wet scrubbing systems. Since the liquid itself and how it is administered are the greatest contributors to the performance of the scrubber, the liquid circuit must get special attention. Experience has shown that pumps are reliable devices when they receive proper maintenance. The pump supplier's maintenance procedures should be followed to obtain the best performance from the pump(s). How the liquid is administered into the scrubber, however, varies by design. Therein lie many problems. If the scrubber removal efficiency is a problem, the first point to look at is where the liquid meets the gas, then work backward to the pump. Most problems can be localized to improper liquid rate or distribution at the point of injection.

For example, in spray tower scrubbers, a primary cause of lower scrubber efficiency is nozzle plugging or failure. If a nozzle plugs or breaks, the spray pattern is no longer of the type or shape required by the designer. This change may adversely affect performance. Sudden pressure drop or flow changes as revealed by a system datalogger may indicate the mechanical loss of one or more nozzles. A gradual decline in flow or increase in pressure drop could indicate a slow plugging of the nozzle or its supply header. Often, the tower must be shut down and the nozzles be individually inspected. Spare nozzles are a good insurance policy for spray scrubbers.

In packed towers with spray type liquid distributors, the same holds true. A more common problem, however, is an insufficient number of nozzles, that is, poor liquid distribution. You may see this if you observe the top of the packed bed and see spray-pattern-induced discoloration or packing solids buildup. The remedy is to measure the existing spray pattern and design a new one with fully overlapping coverage. Sometimes, wider spray-angle pattern nozzles are all that are needed.

A more common occurrence in spray distributor type packed towers is that too much packing has been installed in the tower. The distance between the packing and the spray nozzle is too low to allow the spray pattern to form. Another problem is the use of nozzles (so-called pig-tail type) that produce a rose petal pattern near the nozzle and a more uniform pattern away from the nozzle. These nozzles work well at the intended nozzle-to-packing distance but can be a problem if that distance is reduced. If the packing must be up close (say, to get greater removal efficiency), check with the nozzle vendor for a replacement selection that produces a full cone pattern in the nozzle or in proximity to the nozzle discharge.

On Venturi scrubbers, the liquid headers are usually open pipes and plugging is the only concern. On annular types that use a center pipe, the pipe must be centered on the apex of the cone. Typically, one third of the liquid flow goes to the center pipe and one third goes to the tangential or wall headers.

On fluidized bed or ebulating bed scrubbers, the headers are typically horizontal with open holes drilled horizontally through both walls. These headers are usually "dead-ended", and solids can build up in the far holes. These headers are removed and cleaned to correct the problem. Care must be taken to replace the headers so that the liquid discharges horizontally, not vertically, on most designs.

Flow control valves (especially bleed control valves that open and close) are a common source of hard-to-diagnose problems. They usually create an intermittent or randomly occurring problem. These can sometimes be taken out of the control loop and a 4-20 ma signal injected from a suitable instrument to "fool" the valve into operation. Usually, the positioner for such a valve will not permit full travel and the valve either does not open or does not close fully. Limit switches are commonly used to reverse the action on these devices, and a bit of electronic troubleshooting using a VOM can isolate the problem. Such testing must, of course, be conducted in compliance with the safety and operating constraints of the system. Most often, such tests are conducted when the system is down.

Scaling of recirculation piping is one of the most common problems. If chemical washing is not effective, consider changing the piping design to allow simplified removal of problem-prone sections. Some scrubber chemical suppliers offer additives that can sequester the type of scale-forming solids and allow them to be flushed away. These companies, to date, include Nalco Chemical, Betz Dearborn, Quaker Chemical, and many others. The use of these chemicals must be compatible with your water treatment method or sewer regulations.

Instrumentation Issues

Modern-day instrumentation includes built-in diagnostics and troubleshooting methods. Many plants have in-house instrumentation personnel who add their considerable experience to that provided by the vendor. It has been found that most instrumentation problems that occur on a previously properly operating system involve the simple opening of circuits. This could be caused by loose or corroded wiring or by the failure of circuit boards in the respective controllers.

For the former, the VOM and 4-20 ma loop calibrator are useful tools in diagnosing problems. The VOM can help you find open circuits, and the loop calibrator can permit the injection of a known signal so you can evaluate the response.

As mentioned earlier, pFI probes and controllers typically use a temperature compensation circuit that sometimes opens. Usually, a wiring connection opens, but sometimes it is in the probe itself. Not only should the probe be calibrated but its measured temperature should be confirmed. This function is usually a menu item on the controller panel. On older systems, you must check for an open circuit in the temperature compensation portion of the wiring. In five-wire probes, these wires are usually delineated in the operating manual and can be traced easily.

The loop calibrator can be used to check operation after the device is separated from the circuit. Calibrators vary, so one must check the documentation associated with the device. A common problem can be revealed when one investigates any recent instrumentation changes that may have been made. Sometimes, a ground loop is accidentally formed, which provides a false reference voltage for one or more controllers. This is quite common with level control systems that are immersed in conducting solutions (such as resonant frequency or capacitance types). The remedy sometimes involves the installation of a grounded reference electrode in the sump.

Instrumentation vendors want their equipment to work properly. Most have toll-free technical assistance numbers. It is a good idea to list these numbers along with the model and serial number of the system's instrumentation in the front of any operating manual. It is also suggested to have spare controllers or at least output cards for the key system control units. Often, local welding or voltage spikes can accidentally destroy output cards. Isolated output buses are also available from instrument vendors to fix chronic problems regarding spurious signals.

If your system uses draft control via a draft sensor (load cell, etc.), the loop calibrator can be used to set this device up prior to startup. Often, the control does not range from 4 to 20 ma. Usually, zero draft is 12 ma. You need to contact the controller vendor to determine the set-up sequence. The common method is to calibrate the controller for the maximum draft (negative reading) at 20 ma and enter a positive pressure at 4 ma. The SPAN of the controller is used to provide smooth transition through zero pressure.

Many controllers sample (i.e., read a signal) at a rate different from their output control. This is sometimes called RPM, or resets per minute. Basically, a reading from the sensor is followed by an output signal to the controlled device (say, a flow control valve, draft control damper, etc.). A controller that behaves erratically with its output signal resulting in control element hunting or overheating can often be dampened by decreasing the RPM set point at the controller. The controller simply waits longer between sampling sessions. Most scrubbers do not need constant correction. Once every 20-30 s is usually sufficient. Exceptions would be processes that exhibit sudden puffs or surges. In those cases, more frequent resets would be suggested.

Draft instrumentation is typically accommodated using Dwyer Magnehelic gauges or similar devices. These Bourdon tube type or diaphragm type devices can fill with condensate from the scrubber since most scrubbers saturate the gas stream. If this occurs, the gauge will read both the actual pressure and the head of liquid in the sample line. Sample lines should come out of the pressure tap, then up much like a P trap seal, then down below the gauge. A drain tee and valve at the low point usually allow the condensate to be drained. The sample tube would then go back up to the gauge. These problems are very evident on systems used in cold climates. Periodic draining of the line is required.

If compressed air is used to blow back the sample line, do it after disconnecting the gauge; otherwise, you could damage or even destroy the internal mechanism of the gauge.

Keeping a record of your instrumentation service events can be very handy in determining where your maintenance dollars are being spent. Investing in an ultrasonic probe-cleaning device, for example, may mitigate cleaning pH probes daily. Continual scaling of level controls could lead you to changing to a noncontact type control. The service record will help reveal these possible money-saving opportunities for you.

 
<<   CONTENTS   >>

Related topics