2019 E-zine Articles

Version:1.0 StartHTML:000000244 EndHTML:000028953 StartFragment:000002051 EndFragment:000028921 StartSelection:000002051 EndSelection:000028921 SourceURL:http://www.spitzerandboyes.com/wp-admin/post.php?post=3593&action=edit&message=10

January 2019

Applying Paddlewheel Flowmeters

By Walt Boyes

As more plants try to improve production quality, engineers find more locations where they need to monitor flow. Sometimes these installations are critically important, other times they provide backup information. Very often, these locations present a difficult challenge to installing a flowmeter.

When you look at an application like this, usually cost is a big factor. Cost is divided into two parts, of course: cost of purchase and cost of installation. An inexpensive flowmeter can cost a lot to put in, and an expensive flowmeter can be relatively inexpensive to install.

Generally, if the application isn’t “mission critical,” and the flow is a reasonably clean liquid, you can use one of the ubiquitous paddlewheel flowmeters that are on the market. There are at numerous manufacturers of these devices, and they come in two basic configurations. The most common configuration is an “insertion” device that fits into a standardized tee fitting, or, in larger sizes, an insertion fitting like a saddle or welding fitting. There are also paddlewheel flowmeters that are enclosed in a meter body, usually for 1″ or smaller size lines.

More next month on applications.

This article originally appeared in Flow Control magazine (May 2002) at www.flowcontrolnetwork.com.

Boiler Efficiency Improvements Pay Off

By David W Spitzer

Boiler efficiency is a big deal because large industrial and utility boilers can consume in excess of US$100 million per year of fuel.  Therefore, improving boiler efficiency by only 1 percent can represent fuel savings of approximately US$1 million per year while also being more environmentally benign (greener).  Similar fuel savings in smaller boilers will likewise produce significant energy savings relative to their fuel costs.  Therefore, it is important to monitor boiler efficiency (if only periodically) to maintain efficient operation on a consistent basis and identify areas for improvement.

Determining the combustion efficiency of a boiler would seem to be a straightforward task that can be performed by looking up the efficiency in a table for a given fuel based on the excess oxygen in the stack and the stack temperature rise above ambient conditions.  It should be noted that this technique to determine boiler combustion efficiency does not require any flow measurements.

This approach can work well when the applicable temperature and oxygen measurements are available (even if only on a temporary basis) and the boiler uses a single fuel such as natural gas or a specific grade of fuel oil.  In practice, this approach is applicable to a large number of relatively small boilers that are used for generating steam in small industrial plants and in building/facility heating systems.  In addition, tables are available or can be generated for other consistent fuels such as coal and wood with stable levels of moisture.

More next month…

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Boiler Fuel Cost

By David W Spitzer

What is the approximate annual fuel cost to produce 100,000 lb/hr of steam all year?

  1. US$1 million
  2. US$3 million
  3. US$5 million
  4. US$7 million

Commentary

Much more detailed information is needed to perform this calculation accurately, but a rough estimate can be obtained by making appropriate assumptions such as the boiler efficiency (80 percent) and the cost of natural gas (US$5.00/million British thermal units [mmBtu]) while recognizing that approximately 1000 Btu of energy is required to boil a pound of water to make saturated steam.  Using these assumptions, the annual energy cost would be approximately US$5.475 million:

(8760 hr/yr) x (100,000 lb/hr) x (1000 Btu/lb) x (US$5.00 / mmBtu) / (0.80)

The above expression yields US$4.475 million per year which is close to Answer C.

Additional Complicating Factors

The assumptions are different for each plant location and will change over time.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

February 2019

Applying Paddlewheel Flowmeters: Meter of All Trades

By Walt Boyes

Paddlewheel flowmeters are used in a wide variety of industries, from irrigation and chemical processing, to air-conditioning and refrigeration. They are used especially in industries where cost to measure the variable is an object. The meters can also be made of a variety of corrosion-resistant materials, including PVDF (Kynar) and various other fluorocarbons. They can even be made out of high-pressure materials. Some of the enclosed paddlewheel meters can be used for very high pressures in hydraulic lines and in manufacturing plants for cutting oil flows. They are regularly used in industrial water treatment and in small municipal water treatment plants. You can find them in just about any boiler operation, in the low temperature lines. They are used in scrubbers and semiconductor applications where very high purity is not an issue.

More next month about insertion paddlewheel flowmeters.

This article originally appeared in Flow Control magazine (May 2002) at www.flowcontrolnetwork.com.

Using Tables to Determine Boiler Combustion Efficiency

By David W Spitzer

The use of tables to determine boiler combustion efficiency for a given fuel based on the excess oxygen and temperature rise above ambient conditions was previously discussed.  However, accurate determination of boiler combustion efficiency using tables may not be possible under other circumstances and operating conditions, especially in larger and more complex boiler operations.

For example, the use of alternate fuels that have heating values inconsistent with those used to calculate the efficiencies in the tables will invalidate the assumptions made to produce the tables, often reducing accuracy to the extent of rendering the tables useless.  Likewise, the heating value of non-purchased fuels and wastes introduced into boilers and incinerators can vary all over the map.  Negative heating values can occur when additional energy is required to vaporize and destroy wastes such as wastewater.

In addition, other streams and processes are often present that can affect the accuracy of the boiler combustion efficiencies cited in the tables.  For example, a significant amount of heat loss can occur because of boiler blowdown that is dependent on the quality of the feedwater (determined by the design and operation of the raw feedwater treatment system), the amount of condensate returned to the boiler, the amount of contaminate introduced into the condensate return and the chemical control of the water in the boiler, among other influences.  Soot blow operations similarly cause heat losses (albeit relatively small) that the tables do not address.

In general, more complicated operations usually invalidate the efficiency calculations used to generate the tables and render them of limited or no use.

More next month…

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Measurements Used to Calculate Boiler Efficiency

By David W Spitzer

Which of the following measurements can be used to calculate boiler efficiency?

  1. Feedwater flow
  2. Blowdown flow
  3. Steam flow
  4. Drum level
  5. Drum pressure

Commentary

The feedwater flow (Answer A), blowdown flow (Answer B) and steam flow (Answer C) can be used to calculate the flow of heat either into or out of the boiler.  As such, they can be used to calculate boiler efficiency.  However, as a practical matter, the blowdown flow is rarely measured because it typically contains a relatively small amount of heat.

The drum level (Answer D) is not indicative of heat flow and would not be used to calculate boiler efficiency.

The drum pressure (Answer E) is not indicative of heat flow but could be used to pressure compensate the steam flow measurement.

Additional Complicating Factors

Additional measurements such as the feedwater temperature, multiple fuel flows and the like might be needed to calculate boiler efficiency of the system at hand.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

March 2019

Applying Paddlewheel Flowmeters: Insertion Flowmeters

By Walt Boyes

All of the insertion-style units rely on the point-velocity theory for operation. That is, they sense velocity at a single point. Depending on the flow profile that velocity is either very close to the average velocity in the line or it isn’t. If it is close, the flowmeter is relatively accurate. If it is not, the meter is relatively inaccurate, depending on how far away from the average velocity point the paddle is sitting. Even an inaccurate meter can be quite repeatable however so it is usually possible to use them for control. Generally, the insertion-style paddlewheels have installed accuracies of around five percent of span, regardless of their stated accuracy. In my experience, installing with extreme care can sometimes provide the stated accuracy, especially in lines under 2″.

More next month about viscous flow.

This article originally appeared in Flow Control magazine (May 2002) at www.flowcontrolnetwork.com.

Using Principles to Calculate Boiler Efficiency

By David W Spitzer

Previously discussed was the use of tables to determine boiler combustion efficiency.  This approach is limited to a given fuel and is based on the excess oxygen and temperature rise above ambient conditions.  Many factors can reduce the accuracy of the tables and often render them useless, so first, principles may have to be applied to calculate boiler efficiency.

Boiler efficiency can be defined as the energy content of the steam produced divided by the amount of energy consumed to produce the steam.  This might appear to be straightforward, but it is not necessarily an easy task.

Calculating the energy consumed typically requires a fuel flow measurement that (for gaseous fuels) is often compensated for the operating temperature and operating pressure of the gas.  The energy content of multiple fuel streams must be added to obtain the total energy consumed, especially when multiple fuels are fired at the same time.  The energy content of the steam can be measured with a steam flowmeter that is usually compensated for pressure and sometimes temperature (especially for superheated steam).

The calculated boiler efficiency based on actual process measurements (performed manually or continuously) can be used as a gauge to determine how well the boiler is operating.  The start of a decreasing efficiency trend could be used as an early warning to investigate the problem before the boiler becomes inefficient or fails.  To increase the confidence in the steam flow measurement, steam flow measurements should be periodically compared to feedwater flow measurements to ensure that they maintain the same ratio, which is largely dependent on the type of water treatment system and controls.

More next month…

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Boiler Measurement Problems

By David W Spitzer

What is wrong (or not) with the following actual boiler measurements?

Feedwater conductivity setpoint        2,500 microsiemens (µS/cm)

Feedwater conductivity                      2,550 µS/cm

Drum level setpoint                                50 percent

Drum level                                              47 percent

Feedwater flow                                     100 liters per minute

Steam flow                                                6 metric tons (MT)/hr

Commentary

The feedwater conductivity and drum level are near their respective setpoints and are likely operating properly.  Water chemistry and historical charts should be examined to confirm this presumption.

The density of cold water is approximately 1 gram per cubic centimeter (g/cc), so the operating density of hot feedwater would be lower (say 0.9 g/cc), and the feedwater flow would be approximately 0.9 x 100 x 60, or 5,400 kilograms per hour (5.4 MT/h).

By observation, the steam flow is higher than the feedwater flow.  This cannot occur, so something is incorrect.  Flowmeters that measure steam are typically less reliable than flowmeters that measure hot water, so (absent other information) the operation of the steam flowmeter should be investigated.

Additional Complicating Factors

Further investigation should be undertaken to determine the expected steam to feedwater flow ratio typical of the installed feedwater treatment system.  Knowing the expected ratio will make it easier to know when the investigation reveals the problem.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

April 2019

Applying Paddlewheel Flowmeters: Viscous Flow)

By Walt Boyes

One of the nice things about paddlewheel flowmeters is that they can, under specific conditions, handle viscous fluids quite well. In order to apply them, you have to understand the concept of the Reynolds number. The Reynolds number is a dimensionless value that describes the nature of flow in pipes. It is affected by the pipe diameter, velocity in the line, fluid density and the viscosity of the fluid in the pipe.

The key here is the diameter and the velocity. If you can keep the Reynolds number high enough at the lowest flow and maintain sufficient upstream straight run, the fluid velocity profile will be fully-developed, regardless of changes in viscosity. For paddlewheel flowmeters, that point is a Reynolds number greater than about 4,800. Remember, though, that the Reynolds number must be above that value for the entire flow range or the meter will not be accurate, and may not even be repeatable, especially at the lower flows.

This article originally appeared in Flow Control magazine (May 2002) at www.flowcontrolnetwork.com.

Boiler Efficiency Monitoring Addresses and Avoids Problems

By David W Spitzer

The decreasing steam boiler efficiency alluded to in a previous article can be the result of one or more problems in one or more parts of the system.

For example, decreasing boiler efficiency could be caused by a problem in the feedwater treatment system that causes excessive blowdown that creates excessive heat losses, a faulty chemical treatment probe, a leaking blowdown valve, contamination of the condensate return system, a failing fuel flowmeter, a drifting steam flowmeter, by operating the boiler at a low firing rate, and so on.

Monitoring boiler efficiency can provide an early indication that a problem may exist that should be investigated.  Regularly checking the ratio of the steam flow to feedwater flow helps provide operators with more confidence in the performance of the steam flowmeter.

The existence of two (or more) problems often leads to the symptoms of one or both of the problems to be partially or fully masked.  Therefore, it is suggested that an effort be made to regularly (even if periodically) monitor boiler efficiency to enable problems to be resolved almost immediately after their symptoms first appear.  Allowing problems to accumulate makes their identification more difficult and their resolution more frustrating because the resolution of one problem often unmasks the existence of another problem or problems.

My experience has been that locating and addressing one problem as soon as it occurs is much easier than locating and addressing two more problems with symptoms that often mask each other and potential additional problems.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Boiler Efficiency Decrease

By David W Spitzer

Which of the following could cause boiler efficiency to decrease?

  1. Blowdown valve is leaking
  2. Conductivity setpoint is too low
  3. Conductivity sensor measures high
  4. Steam flow measurement is low
  5. Feedwater flow measurement is nonfunctional

Commentary

The steam and feedwater flow measurements (Answers D and E) can be used to calculate boiler efficiency, but do not affect the actual efficiency of the operating boiler.

A leaky blowdown valve (Answer A) will continuously cause heat to be lost.  Boiler efficiency will decrease when the flow rate through the leaky blowdown valve exceeds the average blowdown flow rate.

Setting the conductivity setpoint too low (Answer B) will cause the blowdown valve to open wider, causing more heat losses than would occur had the setpoint been set correctly (higher), which decreases boiler efficiency.  A conductivity sensor (Answer C) that measures high will effectively make the system operate at a lower conductivity and have essentially the same effect.

Additional Complicating Factors

Numerous other problems can decrease boiler efficiency.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

May 2019

Applying Paddlewheel Flowmeters: Operation

By Walt Boyes

Paddlewheel flowmeters are inexpensive and it is easy to do maintenance on them. Most paddlewheel flowmeters have virtually no salvage value, so it is easier to just buy a new insertion probe, and toss the broken one. Some meters are now being supplied with better shafting and bearings, and these tend to operate longer than those with older style rotor assemblies. One of the maintenance issues associated with paddlewheel flowmeters is that their performance tends to degrade as the rotor and bearings wear. Mostly, this shows up in poorer low flow performance.

Paddlewheels have varied specifications for low flow performance. Some meters claim very low flow performance (0.5 fps), while others claim one fps as the low flow cutoff. It is not a good idea to use an insertion type paddlewheel where the flow regularly falls below one foot per second velocity. If the line size is less than 1″, however, the closed-body paddlewheels tend to operate reliably at much lower velocities than do insertion probe style paddlewheels.

All paddlewheel meters are susceptible to breakage from objects in the pipe and hydraulic shock, so it is wise to locate them where these issues do not apply. Generally speaking, it is not a good idea to use them for start-stop applications unless they are behind a slow-acting valve.

Paddlewheel meters, when applied properly, provide an inexpensive means of measuring flow in non-mission-critical locations.

This article originally appeared in Flow Control magazine (May 2002) at www.flowcontrolnetwork.com.

Backward Problem Solving Can Produce Uncertainty

By David W Spitzer

Legal work is interesting in the (twisted) sense that diametrically opposite approaches to an investigation can be taken to determine the cause of a problem.  There is some art to knowing when to take which approach, but the selected approach is also dependent on the skill set of the person(s) involved.  Sometimes it does not matter, but quite often the results can be strikingly different.

Confidentiality does not allow me to divulge details of actual investigations, so I will invent an example to illustrate the point.  Suppose the problem is to discover what caused a round splat on a downtown sidewalk that is about 0.5 meters in diameter.  The liquid part of this artifact is clear with some yellow coloring and brown/white chips throughout.  Given its location, shape, color and consistency, one might speculate that the splat was made by a brown egg falling from above.  Chemical testing can confirm that an egg was involved.

The roundness of the splat indicates that the egg likely fell from directly above.  But which floor did it fall from?  Without elaborating, different tests with different quality brown eggs can be performed to determine which quality egg dropped from which floor exhibits the same splat as the original splat.  But what if the egg did not fall from a rest position but instead was thrown up or down?  This question adds a certain amount of uncertainty.

The point is that starting with an artifact and working the problem backward can create considerable uncertainty. Nonetheless, working the problem backward is often the best — and sometimes the only — way to work certain problems.  An example might be determining the nature of dinosaurs when you only have artifacts to study.

More next month…

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Magnetic Flowmeter Accuracy Statements

By David W Spitzer

Which of the following forms describe how the accuracy of a magnetic flowmeter would typically be expressed?

  1. Percent of rate
  2. Percent of full scale
  3. Percent of measured value
  4. Percent of calibrated span
  5. Percent of span

Commentary

The accuracy of magnetic flowmeters is typically expressed as a percentage of rate, (Answer A) which is the same as a percentage of measured value (Answer C).  However, the additional accuracy associated with the accuracy of the analog output is typically expressed as a percentage of full scale (Answer B) that is the same as a percentage of span (Answer E).

Additional Complicating Factors

The measurement accuracy of the analog output is the magnetic flowmeter accuracy plus the analog output accuracy expressed as a percentage of rate.  In addition, accuracy is degraded at low flow rates.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

June 2019

Positive Displacement Flowmeters

By Walt Boyes

Sometimes you have to be very exact in your flow measurement. Sometimes you are being charged, or charging someone else for your product. Sometimes you need to have a carefully controlled amount metered into a process.

Sometimes you need a positive displacement flowmeter.

Positive displacement meters, often called PD meters, have been around for quite a while. They are, in essence, an automated bucket. One bucket in equals one bucket out. Pretty simple. PD meters have their own applications knowledge base and have their share of pitfalls, also.

There are several basic designs of PD meters. PD meter design effectively mirrors positive displacement pump design. For every type of positive displacement pump, there is a PD meter based on that type. There are reciprocating and rotary piston meters, gear meters, nutating disc meters and diaphragm meters. Each of these meters is designed to provide a volume on the inlet of the meter that fills up, and then move it to the outlet of the meter, allowing the inlet volume to refill.

More next month about applications.

This article originally appeared in Flow Control magazine (March 2002) at www.flowcontrolnetwork.com.

Choosing Between Forward and Backward Problem Solving

By David W Spitzer

The previous article included discussion of how working a problem backward from an artifact to find the cause of a problem can create considerable uncertainty.  Nonetheless, this approach is sometimes the best — and often the only — way to obtain information (such as when studying dinosaurs).

Now let us consider the opposite approach of working my (previously) invented problem of determining the cause of the splat on the sidewalk. Suppose that you found witnesses who were in a second-floor office located directly above the splat who saw a person drop a brown egg out of the window a few minutes before the splat was found by a passerby.  Now we can work the problem forward and reasonably expect that the egg should cause a splat on the sidewalk immediately below that would be consistent with the observed splat.  This can be confirmed by dropping other similar eggs and comparing splats.

The point is that working the problem forward from known events generally creates less uncertainty than working the problem backwards from an artifact, understanding that working the problem backward is often the best and only method to address certain problems.

Knowing which approach is appropriate for the problem at hand — or if combinations of the two are warranted — is more an art rather than science.  Interestingly, we often find ourselves embarking on one of these approaches without formally thinking about it.

More next month…

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Differential Pressure Flowmeter Accuracy Statements

By David W Spitzer

Which of the following forms describe how the accuracy of a differential pressure flowmeter would typically be expressed?

  1. Percent of rate
  2. Percent of full scale
  3. Percent of measured value
  4. Percent of calibrated span
  5. Percent of span

Commentary

This question is one that would be asked in a plant but it is not precise enough to elicit a definitive answer.  The primary element (such as an orifice plate) has an accuracy expressed as a percentage of rate (Answer A) or measured value (Answer C).  However, the accuracy of its differential pressure transmitter can be expressed as a percentage of full scale (Answer B), calibrated span (Answer D), span (Answer E), upper range limit or a combination thereof.

Additional Complicating Factors

The accuracy of a differential pressure measurement system includes the performance of the primary element and its transmitter.  The transmitter accuracy expressed as a percentage of rate increases as flow decreases, so the combined accuracy expressed as a percentage of rate will be different at different flow rates.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

July 2019

Positive Displacement Flowmeters: Common Applications

By Walt Boyes

You would think that PD meters could be used for nearly everything, but you would be wrong. They cannot be used with fluids with particulate solids. They are useless for wastewater. They can only be used for clean fluids that do not bind the action of the meter.

The single most common application for PD meters is water utility distribution metering. Millions of PD meters, mostly rotary piston or nutating disc type, are manufactured every year for the municipal utility market worldwide. Another form of near positive displacement meter, the multijet, is also used for this purpose.

Next most common is oil and gas metering, both industrial/commercial and residential. Positive displacement meters are found in almost every gasoline station, LNG tank dispenser and home oil burner.

Positive displacement meters, with contact closure pulse outputs, are also used in the control of boilers and cooling towers. There are many other devices that could also be used, but the pulse meter, as these devices are called, is the most common. This is because they are a simple variation on the water utility meter. They are very inexpensive because water utility meters are made in enormous quantities.

More on a cooling tower application next month.

This article originally appeared in Flow Control magazine (March 2002) at www.flowcontrolnetwork.com.

Different Perspectives Often Needed for Effective Problem Solving

By David W Spitzer

In previous articles, I discussed how working a problem backward from an artifact to find its cause can create considerable uncertainty.  Nonetheless, this approach is sometimes the best and often the only way to find a cause (such as when studying dinosaurs).  In contrast, working the problem forward (when possible) tends to create less uncertainty.

Being confronted with how to approach a problem actually occurred in a recent legal case (of course, not with eggs as previously described).  In this case, various experts who were ostensibly knowledgeable in their areas of expertise examined the artifacts to determine the cause of the problem.  However, the perspectives of these experts were limited to their respective fields.

You might say that, “to a hammer, everything is a nail.” As a child, I remember hearing that, “if you go to a surgeon, you get surgery.”  Given an artifact, a mechanical engineer might examine its shape and strength, a metallurgist might determine its composition, a corrosion specialist might determine its corrosive properties and whether evidence of corrosion was found, a civil engineer might determine if it was adequately supported, an electrical engineer might determine if it was wired properly, a piping engineer might determine if it was piped properly, and so on.

In this particular case, examining the artifact created a significant amount of uncertainty as to what failed and why.  Starting with the known operating conditions and performing calculations to work the problem forward revealed that operating conditions existed that would cause damage similar to that observed in the actual artifacts.

This is not to say that instrumentation engineers know all.  It often takes an open mind and multiple experts to examine problems from different perspectives to determine what really happened and resolve the problem in future operation.  This same approach can be taken at the local plant internally and (when appropriate) by hiring the proper expert with the proper skill set at the proper time to examine the proper problem(s).

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Weep and Vent Hole Orientation in Horizontal Pipe

By David W Spitzer

How should weep holes and vent holes of an orifice plate be oriented in a horizontal pipe?

  1. Top of pipe
  2. Three o’clock position
  3. Bottom of pipe
  4. Nine o’clock position
  5. Side of the pipe

Commentary

The idea behind weep holes and vent holes is to remove the fluid in the nonflowing state from the flow measurement system.

For gas service, the weep hole should be located at the bottom of the pipe (Answer D) to allow liquids that may be present upstream of the orifice plate to pass downstream.

For liquid service, the vent hole should be located at the top of the pipe (Answer A) to allow noncondensable gases that may be present upstream of the orifice plate to pass downstream.

Additional Complicating Factors

Orifice calculations should compensate for the flow that passes through the weep holes and vent holes during normal flowmeter operation.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

August 2019

Positive Displacement Flowmeters: Cooling Tower Application

By Walt Boyes

A typical application for PD meters is cooling tower makeup water control. Typically, in a decent sized cooling tower in a manufacturing facility or process plant, a two-inch bronze PD meter with a pulse output is used, and the meter is located close to the spout of the makeup water pipe, downstream from a solenoid valve. I picked this application because it illustrates many of the difficulties with applying a PD meter, all in one common application.

Water is incompressible. When it travels down an air filled pipeline, the leading edge of the water surge is a pressure wave, sometimes called water hammer. Typically, water hammer has a pressure surge of at least five times the static head in the pipe. So, if the pressure in the pipeline is typical plant water pressure, 80 psig or more, the water hammer will have the pressure of 400 psig or more.

In this application, the solenoid valve slaps open, sending a pressure wave of water toward the flowmeter, which is standing empty. The shock wave hits. The meter internals, if they aren’t shattered from the blow, begin to spin. The meter moves extremely rapidly, as the metering chambers fill up with air, water, and then the air is pushed out of the meter by the velocity of the flow. The meter sends a stream of pulses to the process controller.

What’s wrong with this picture?

This article originally appeared in Flow Control magazine (March 2002) at www.flowcontrolnetwork.com.

Spotting Fume Flow Measurement Problems

By David W Spitzer

The time it takes for people to recognize an error or oversight can be small (think about a hand exposed to fire) or take decades.  The consequences of the former are usually learned in a short period of time, often shared with others and hopefully last for a lifetime.   The latter often entail subtleties that become apparent serendipitously or when the problem appears again many years later.

Some years ago, the plant used an existing insertion Pitot tube to measure the fume flow to the incinerator.  The flow of fumes not only provided cooling to the fume nozzles, but also represented a significant mass flow in the incinerator.  Therefore, the fume flow was measured to assure the operator that the fume nozzles were being cooled, and to provide information so the operator could operate the auxiliary air fan when insufficient fumes were available to cool the fume nozzles.

The fume flow measurement was not used for control and the process was straightforward — the fumes or auxiliary air were routed to the fume nozzles — so the existing fume flowmeter measured the total fume flow.  It typically measured approximately zero, 30 or 60 percent of full scale to reflect the number of operating blowers.

When the incinerator was upgraded, the fumes were partially used in place of a portion of the combustion air.  This change provided superior destruction of wastes in addition to reducing fuel consumption and increasing incineration capacity.

The downside was that the fume piping was more complex and the fume flow to the fume nozzles needed to be continuously maintained and controlled.  In addition, the fume flow could replace up to 80 percent of the combustion air and would drop to near zero flow approximately 5 seconds after the fume blowers turned off to either stop production or trip the unit (which could happen at any time).

More next month…

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Vertical Pipe Weep and Vent Hole Orientation

By David W Spitzer

How should weep holes and vent holes of an orifice plate be oriented in a vertical pipe?

  1. Top of pipe
  2. Three o’clock position
  3. Bottom of pipe
  4. Nine o’clock position
  5. Side of the pipe

Commentary

A recent Quiz Corner asked the same question and presented the same answers with regard to a horizontal pipe installation where the answers had specific orientations for liquid and gas flows.

If you have to give an answer to this question, Answer E would be the best answer because the vent hole and weep holes are located near the side of the pipe.

However, vent holes and weep holes are not needed when the orifice bore itself will allow liquids (in gas service) and noncondensable gases (in liquid service) to pass downstream, making this question moot.

Additional Complicating Factors

None!  Why complicate the situation?

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

September 2019

Positive Displacement Flowmeters: The Big Four (Part 1)

By Walt Boyes

Last month, I described an application where a solenoid valve slaps open sending a pressure wave of water toward the flowmeter, which is standing empty. The shock wave hits. The meter internals, if they aren’t shattered from the blow, begin to spin. The meter moves extremely rapidly, as the metering chambers fill up with air, water, and then the air is pushed out of the meter by the velocity of the flow. The meter sends a stream of pulses to the process controller.

What’s wrong with this picture?

First, the meter is left dry, and is subjected to shock and water hammer every time it turns on. This produces extensive excessive wear on the meter internals.

Second, the meter turns extremely fast for the first second or so as the water pressure equalizes in the meter body. A two-inch bronze meter is designed with a maximum flow rate of just over 100 gallons per minute of flow. The meter, if overdriven, can be subjected to massive wear or even destroyed. Flow out of an open two-inch diameter pipe can easily exceed 100 GPM with 80 psig line pressure behind it.

More on the Big 4 next month.

This article originally appeared in Flow Control magazine (March 2002) at www.flowcontrolnetwork.com.

Incinerator Fume Flowmeter Replacement

By David W Spitzer

My previous article described the existing installation and operation of a fume flowmeter for an incinerator.  The new installation is more complex and uses fumes to replace up to 80 percent of the combustion air, but the fume flow can drop to near zero flow in approximately 5 seconds after the unit shuts down or trips.

The existing Pitot tube flowmeter measured reliably over the years, but it had its negatives.  In particular, the Pitot tube is an inferential flowmeter; its low-range, differential pressure transmitter was difficult to calibrate, it drifted (especially when subjected to sunlight), and its squared output degraded accuracy at the low flow rates that needed to be measured in the new process.

The new process diverted most of the fumes away from the fume nozzles to the burner, so the new combustion control strategy required measurement of its flow to achieve stable combustion.  In addition, the measurement of fume flow to the fume nozzles was not feasible, so measurement of the fume flow to the burner was also utilized to calculate the fume flow to the fume nozzles (which was monitored and controlled).  In other words, this measurement was important because it played a role in keeping the fume nozzles from overheating as well as account for the equivalent of up to 80 percent of the combustion air to the burner.

An insertion thermal flowmeter was selected for this application because it was linear and able to measure mass over a wide flow range of flows.  The new flowmeter worked well for a week before it started to measure incorrectly.  The sensor was removed, cleaned and returned to service.  The same symptom returned a few days later, so the new flowmeter was replaced with an insertion Pitot tube similar to the existing fume flowmeter that had been in service for years.

More next month…

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Coriolis Mass Flowmeter Sizing

By David W Spitzer

The plant manager wants to relocate a 2-inch Coriolis mass flowmeter measuring zero to 200 kilograms/minute of water that is no longer needed to another part of the plant to measure the same flow of chocolate.  Will this work?  If not, what problems do you anticipate?

Commentary

This could be a tricky question because it involves technical analysis and (potentially) politics.

From a technical perspective, the chocolate is much more viscous than water, so the pressure drop across the flowmeter will be much greater than that for water flow.  The pressure drop across the flowmeter can be obtained using the Coriolis manufacturer’s pressure drop curves or software.  If the pumping system cannot generate sufficient pressure to achieve full flow, a larger Coriolis mass flowmeter may be required.  Given the much higher viscosity of chocolate, it is likely that at least a 3-inch Coriolis mass flowmeter will be needed.

Additional Complicating Factors

Presuming that a larger flowmeter is necessary and despite your best efforts to explain the technical details, the plant manager may nonetheless order you to install the 2-inch Coriolis flowmeter that will not work properly.  If this occurs, I suggest documenting your actions and have the larger flowmeter and its associated piping available on-site (or at least nearby).  By the way, we used to call this an “object lesson.”

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

October 2019

Positive Displacement Flowmeters: The Big Four (Part 2)

By Walt Boyes

Continuing discussion describing problems associated with the application discussed last month…

Third, the meter depends for its accuracy on all of the chambers in the meter being filled completely with fluid. That is the “positive displacement” the meter is named for. In this application, the meter’s chambers may retain some trapped air. Since some of the volume in the measuring chamber is then air, the meter is not accurate.

Fourth, in this application, the inlet water to the meter is sometimes a recirculation line from the cooling tower’s sump. In that case, there are usually sediments, sand or other solids, like scale and calcium carbonate deposits that have flaked off the inside of the pipes or the tower itself, present in the water. These particulates are extremely abrasive and immediately increase the wear on the meter. This increases the leakage through the meter, and therefore the meter’s accuracy decreases.

So, in one application, we have seen four of the worst things you can do to a PD meter: shock, overdriving, air or gas entrainment and solids in the fluid to the meter. These are the Big Four to avoid.

There are other bad things you can do, including changes in viscosity, temperature and flow profile, which will seriously affect the accuracy of a PD meter. Accuracy is the most important feature of a PD meter. Anything you do to reduce the meter’s measured accuracy will detract from the application, so watch out for the Big Four.

This article originally appeared in Flow Control magazine (March 2002) at www.flowcontrolnetwork.com.

insertion Flowmeter Thermal Probe Orientation

By David W Spitzer

Previous articles described the existing installation and operation of a fume flowmeter for an incinerator.  The new installation was more complex and used fumes to replace up to 80 percent of the combustion air but could drop to near zero flow in approximately 5 seconds.  The new fume flowmeter measurement was used in the combustion control strategy and also in the control strategy to ensure that the fume nozzles did not overheat.

Examination of the new thermal flowmeter revealed encrustation on the thermal sensor.  Discussion with the manufacturer indicated that the encrustation was likely caused by small droplets in the fume that hit the heater and were subsequently evaporated by the heater leaving behind a residue that caused the flowmeter to function improperly.  Although the new insertion Pitot tube flowmeter functioned reliably for years, the plant would have preferred to have had the superior measurement available from the insertion thermal flowmeter.

Recent discussions with a thermal flowmeter manufacturer brought to light another potential cause of the encrustation.  In humid flowing streams, the outside temperature is often lower than the gas temperature, so liquid tends to condense on the pipe wall.  When the thermal probe is mounted on the top of the pipe, any condensate formed at the top of the pipe and its tap can travel down the probe to the sensor, where it can adversely affect the measurement.

For humid gas service, thermal flowmeter manufacturers suggest that insertion thermal probes be oriented such that condensation on the pipe walls cannot travel down the probe to the sensor.  This can be achieved by installing the probe pointing upward from the bottom of the pipe or upward at up to a 45-degree angle from the bottom of the pipe so condensation on the probe will flow away from the sensor.

After almost three decades, I now question whether this detail was the difference between success and failure.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Differential Pressure Across Orifice Plate with Increased Flow

By David W Spitzer

Increasing flow from 100 to 200 liters per minute (lpm) will cause an orifice plate primary flow element to generate:

  1. 25 percent of the differential pressure at 100 lpm
  2. 50 percent of the differential pressure at 100 lpm
  3. The same differential pressure as at 100 lpm
  4. 200 percent of the differential pressure at 100 lpm
  5. 400 percent of the differential pressure at 100 lpm

Commentary

The differential pressure across an orifice plate primary flow element will increase as flow increases, so Answers A, B and C are not correct.

Orifice plate primary flow elements produce a differential pressure that is proportional to the square of the flow rate.  Therefore, doubling the flow will create (200/100)2 or four times the differential pressure (Answer E).

Additional Complicating Factors

The squared output relationship applies to concentric orifice plates operating in the turbulent flow regime.  Conical, eccentric, integral, quadrant and segmental orifice plate designs generally follow this relationship over a limited range of Reynolds numbers.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

November 2019

Application Impossible: Powdered Cement (Part 1)

By Walt Boyes

Control system professionals get these questions all the time. They appear on bulletin boards, in newsgroups, and on control-related mailing lists. They come on the phone and in the mail. The question usually starts, “How do I …”

Nearly, every one of them deals with a difficult-to-impossible application.

One recent plea asked: “How do I measure the flow of powdered cement in a pneumatic conveyor tube?”

For those of you who have never seen this sort of flow application, here are some of the particulars. The conveyor tube is generally round, usually between 3″ and 6″ diameter, and is under vacuum. The product is sucked along the tube by the vacuum, and to keep the product from bridging, blocking and rat-holing in the conveyor tubes, the ratio of product to volume is kept low, usually less than 20 percent.

Since the airflow is relatively constant and only changes if there is a blockage, you cannot just measure the flow of air and say that it represents the flow of product.

So, you need to figure out how to do the mass flow.

More next month.

This article originally appeared in Flow Control magazine (February 2002) at www.flowcontrolnetwork.com.

Turbulent Flow Profile Versus High Velocity

By David W Spitzer

A recently published article addressed the K factor of a particular class of flowmeters in liquid service by stating that, “Should the flow profile of a pipeline transition from turbulent (high velocities) to a laminar flow regime (low profile), flow measurement technologies may experience a change in K factor that can result in low resolution.”

Note that the author associates a turbulent flow profile with high velocity.  While related (via Reynolds number), they are different.  A turbulent flow profile relates to a profile typical of operating at a Reynolds number that is more than approximately 4,000.  Viscous liquids typically operate completely in the laminar flow regime and often do not exhibit a turbulent flow profile at high velocities.  Interestingly, viscosity (used to calculate the Reynolds number) was not mentioned in the article.

The author associates a turbulent flow profile with a low profile.  What is a “low profile?” and how do changes in K factor result in low resolution?

The article continued by stating to, “Always consult the with the meter manufacturer if the Reynolds number will be between 2,100 (laminar) and 4,000 (turbulent flow).”  Some flowmeter technologies do exhibit nonlinearity between Reynolds numbers of (say) 2,100 and 4,000, but I should not have to “always” consult meter manufacturers because only a few technologies exhibit this nonlinearity.  Interestingly, this phenomenon applied to only one technology mentioned in the article.

What was the author trying to say?  I am not sure, but I suggest: Be careful because some flowmeters can operate non-linearly in the transitional flow regime.

This situation could have been avoided if the article were either written by a subject matter expert as a “work for hire” or reviewed by a subject matter expert prior to publication.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Orifice Plate Designs

By David W Spitzer

Which of the following are orifice plate designs?

  1. Concentric
  2. Conical
  3. Eccentric
  4. Integral
  5. Quadrant
  6. Segmental

Commentary

The most common design is the thin concentric orifice plate (Answer A), which operates in the turbulent flow regime.  Eccentric (Answer C) and segmental (Answer F) orifice plates can handle similar applications for fluids with solids in liquid or liquids in gas.

Other designs were developed to handle applications in which the previously mentioned designs could not be used.  Conical (Answer B) and quadrant (Answer E) orifice plates can operate at relatively low Reynolds numbers.  Integral orifice plates (Answer D) are used to measure small flow rates.

Additional Complicating Factors

Most of the research on differential pressure flowmeters over the last 90 years or so was performed on thin, concentric orifice plates.  Therefore, the flow coefficients and equations associated with concentric orifice plates are quite good.  Other orifice plate designs will likely result in inferior performance as compared to a concentric orifice plate.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

December 2019

Application Impossible: Powdered Cement (Part 2)

By Walt Boyes

The problem with measuring the flow of powdered cement in a pneumatic conveyor tube is that the mass per unit volume is small. You cannot use any sort of densitometer to give you a density value to multiply flow by. The differential density is too small to see with any density measuring device, except an optical particle counter which tend to get quite expensive in these applications.

My point being, there are applications in flow control (and in motion, reactor, level, and other types of sensor-based control) where there just isn’t a way to do it.

There isn’t unless you change the rules, of course.

The proper solution to the application for the pneumatic conveyor is to break it up, put a chute in the middle of the conveyor system, and use a rotating paddle solids flowmeter.  (There is a tradename product called Bindicator which these things are generally referred to as, whether they are made by Venture Measurement or not.) A plugged chute detector switch is used for backup. However, you can’t just stick a flowmeter in the line and expect it to work.

Another application month.

This article originally appeared in Flow Control magazine (February 2002) at www.flowcontrolnetwork.com.

Learn From Mistakes and Think Critically

By David W Spitzer

Every now and then, a published article or internet discussion provides insight into what is really going on.  Often these articles entail trying to technically defend the technically indefensible.

For example, every 10 years or so, I am asked to review an article about variable speed drives that selectively applies the Affinity Laws to incorrectly calculate energy savings.  My recommendation is usually against publication, but the author typically modifies the article to make his or her analysis technically correct within a small set of uncommon applications.  The information that I draw from these experiences is that the author may be effective in his or her work in the field, but does not really understand the nuances.

One can learn a lot by thinking something through, but often more is learned by making mistakes and observing the consequences.  When possible, it is better to learn from someone else’s mistakes instead of your own mistakes.  However, humans generally do not like to publicize their mistakes (even if it will help others), so opportunities to learn from others’ mistakes can be limited.   The good news is that this seems to be changing because younger workers are more communicative and collaborative than those nearing retirement.

The importance of technical correctness cannot be overemphasized when it comes to technical writing.  A few technical errors will always occur, although some may be caused by inappropriate edits after leaving the author’s desk.  Having written more than 10 books, more than 400 technical articles and numerous whitepapers for our clients, I suggest that you read articles (including mine) skeptically.

One should remember that lessons are more often learned from mistakes than from successes.  However, as you strive to improve plant performance, I suggest that your goal be to know your field well enough so that your mistakes are subtle and do not cause harm before you fix them.  This may require taking calculated risks that you might want to share with others — before and after the fact.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.

Flow Calibration Techniques

By David W Spitzer

Which of the following can be used as a flow calibration technique?

  1. Compare a flowmeter with a flow laboratory standard.
  2. Compare a flowmeter with a master meter.
  3. Compare a flowmeter with a meter under test (MUT).
  4. Verify electronic adjustments (zero/span).
  5. Verify flowmeter dimensions.

Commentary

Comparing a flowmeter with an MUT (Answer C) is incorrect because the flowmeter would be compared with itself.

Comparing the flowmeter with a flow laboratory standard (Answer A) and comparing the flowmeter with a master meter (Answer B) are wet calibration techniques in which the fluid actually flows through the flowmeter.

Verifying electronic adjustments (Answer D) and verifying the flowmeter dimensions (Answer E) are dry calibration techniques that are performed without flowing the fluid through the flowmeter.

Additional Complicating Factors

Wet calibrations are usually more difficult to perform, but they are generally more effective than dry calibrations.  However, dry calibrations are used extensively on some flowmeter technologies, including the calibration of orifice plate flowmeters for the custody transfer of natural gas.

This article originally appeared in Flow Control magazine at www.flowcontrolnetwork.com.