Water/Wastewater eHandbook

Water System Woes

Waylay

2

©2014 The Lubrizol Corporation, all rights reserved. All marks are the property of The Lubrizol Corporation. The Lubrizol Corporation is a Berkshire Hathaway company.GC XXXXXX

See how the details on the inside make all the difference on the outside. Call a piping systems consultant at 1.216.447.7397 or visit corzancpvc.com to learn more.

Choose Corzan HP for Chemical Processing, Semiconductor & Electronics, Water Treatment, Industrial Manufacturing, Mining and Power Generation applications.

Introducing Corzan® HP, creating piping systems with better

CorroSIon rESISTanCEand HIGH-TEMPEraTurE PErforManCEfor transmission of CHEMICal fluIdSall made with lubrIzol TECHnoloGy

to give you MoRe iNSiDe.

™

Corzan HP is a high-pressure, high-temperature, high-impact-strength piping system engineered with Lubrizol’s specially formulated chlorinated polyvinyl chloride (CPVC) compounds that meet the ASTM D1784 certification for 24448 cell class. It’s the only pipe that meets ASTM F441 material classification 4120-06, with a pressure rating 25% higher than standard CPVC at 180ºF (82ºC).

3 3

Table of ContentsProperly Apply Plastic Piping 6Use of such piping requires careful consideration of a number of issues

Succeed at pH Troubleshooting 11Heed some pointers to properly diagnose the cause of a problem

Treat Condensate Water with Activated Carbon 17High-purity coconut-based carbons reduce silica leach, rinsing requirements

Additional Resources 20

4

Jacobi Carbons, Inc. 215-546-3900Columbus, OH, USA www.jacobi.net

Industry Standard Activated Carbons for Wastewater Treatment

Type/Application

Granular, Moderate Activity – TOC / COD RemovalGranular, High Activity – TOC / COD RemovalGranular, Microporous – VOC RemovalPowdered Grade – VOC Removal

Product

AquaSorb™ 80AquaSorb™ 90 AquaSorb™ CS AquaSorb™ CP1

Jacobi Carbons, a worldwide leader in activated carbon technology, manufactures our AquaSorb™ line of activated carbon products to meet the

most challenging wastewater treatment applications, with grades produced from coal, coconut, and proprietary blends of raw materials. Whether the

wastewater treatment challenge is overall organic/toxicity reduction, decreasing chemical oxygen demand (COD), or eliminating specific volatile organic

compounds (VOCs), Jacobi can assist in the selection of the correct grade of carbon, including laboratory testing and analysis.

5

Ad IndexLubrizol 2 www.corzancpvc.com

Jacobi 4 www.jacobi.net

Atlas Copco 10 www.atlascopco.us

Smith & Loveless 16 www.smithandloveless.com

6

FoR yeARS, process plants have relied on non-metallic piping as a cost-effective alternative to stainless, alloy and other expensive metallic piping in applications posing high corrosion rates or requiring stringent cleanliness standards.

Plastic piping affords excellent resistance to attack by many chemicals, including most acids, alkalis and salt solutions. Such piping comes in schedule-40, schedule-80 and other common sizes, with wall thickness usually corresponding to that of steel piping. In addition, some plastic piping, e.g., PVC, is offered with standard dimen-sion ratio (SDR) ratings, which mean the piping system will maintain a more-or-less uniform pressure rating at a specified temperature regardless of pipe diameter.

However, some issues — e.g., thermal movement and other thermal effects, and liquid hammer — demand more attention with plastic piping than with commonly used metallic piping. Most plastic piping materials exhibit a relatively high coefficient of thermal expansion.

Elevated temperatures may seriously affect plastic piping; for some materials, pressure/temperature ratings drop substantially at temperatures above 50°C. So, plastic piping should not be located near steam lines or other hot surfaces. When liquid flow in a piping system stops suddenly (for instance, because of a quick-closing valve), a pressure surge known as liquid hammer (or often water hammer) develops and can easily rupture a plastic piping system.

Plastic piping systems usually require closer support spacing than metallic ones, particularly at elevated tem-peratures. Proper support for plastic piping is essential. As a very rough indication, allowable span is around half that of equivalent metallic piping (same size, schedule, etc.); check the values of allowable span published by the plastic piping’s manufacturer or consult a specialist. Supports and hangers can be clamps, saddles, angles, or other standard types; supports should have broad, smooth bearing surfaces, rather than narrow or localized

Properly Apply Plastic PipingUse of such piping requires careful consideration of a number of issuesBy Amin Almasi, rotating equipment consultant

7

contacts, to minimize the danger of stress concentrations.

Vibrations can damage most plastic-piping systems. Therefore, you must properly assess the piping under dynamic forces and apply mitigation techniques as needed. For instance, plastic piping connected to a large pump might experience high power vibrations, which might necessitate a vibration isolation device. However, such vibration isolators may pose operational or reliability problems, so install them only in special applica-tions where they really are required and no other solution is viable.

Non-metallic piping systems most often rely on a fitting, such as a tee, to provide a branch connection; the fitting usually provides adequate strength to sustain the internal and external pressure of the piping. A branch connection made directly on a pipe weakens the pipe at the location of the opening; unless the wall thickness of the pipe sufficiently exceeds that required to maintain the pressure, you must provide added reinforcement. Assess the amount of reinforcement per applicable codes and specifications.

PLASTIC oPTIoNS

Several types of thermoplastics are available as piping. So, let’s look at those most commonly used.

Polyvinyl chloride (PVC) is stron-ger and more rigid than many other thermoplastics and has relatively high mechanical strength, tensile strength and module of elasticity. It

is both lightweight and low cost, and demands little maintenance. Addi-tionally, various solvent cements or other methods can fuse PVC pipes together to create permanent joints that are virtually impervious to leak-age. Moreover, the material exhibits excellent chemical resistance to a wide range of corrosive liquids. How-ever, PVC requires careful installa-tion to avoid longitudinal cracking and over-belling, and certain liquids such as aromatics and some chlori-nated hydrocarbons can damage it.

The allowable span for PVC piping generally should not exceed around 40%–50% of that of equiva-lent steel piping; the allowable span increases more slowly with diameter compared to steel piping systems. As

a very rough indication, for typical steel piping systems, the allowable spans are 3 m, 5 m and 7 m for 2-in., 6-in. and 10-in. piping, respectively. You also can conservatively estimate the allowable span of steel piping via S = 2 D½ where D is the pipe diameter in inches and the span is in meters. In contrast, for PVC piping, allowable spans are 1.8 m, 2.5 m and 3 m for 2-in., 6-in. and 10-in. pip-ing, respectively; you conservatively can estimate the allowable span of PVC piping via S = 1.4 D⅓, again with diameter in inches and span in meters.

Chlorinated PVC (CPVC) offers higher heat resistance than PVC; because of its excellent corrosion resistance at elevated temperatures,

ReLATeD CoNTeNT oN CHeMICALPRoCeSSING.CoM“Challenge Pipe Specifications,” http://goo.gl/9nzukt “Put CPVC Piping In Its Place,” http://goo.gl/lFPSE2“Test the Mettle of FRP Tanks,” http://goo.gl/YiQCwJ“Master Metering Pumps,” http://goo.gl/bWJCqV“Get Up to Speed on Gear Pumps,” http://goo.gl/ccL4tk“Mind Piping System Mechanics,” http://goo.gl/Ni3pmw“Count on Steam Turbines,” http://goo.gl/3pG7Nl“Make the Most of Centrifugal Pumps,” http://goo.gl/izrF4B“Investigate Integrally Geared Compressors,” http://goo.gl/VOk5F0“Head Off Hassles During Startup,” http://goo.gl/6haqKk“Select the Right Centrifugal Compressor,” http://goo.gl/cmpIim“Take Advantage of Gas Turbines,” http://goo.gl/a4ViHp“Keep Out of Hot Water,” http://goo.gl/lxJKWy“Choose the Correct Turbo-Compressor,” http://goo.gl/2iv63s“Put the Right Spin on Rotating Equipment Revamps,” http://goo.gl/X9RGi5

8

CPVC finds use at temperatures up to 90°C versus the normal limit of 60°C for PVC. However, it is more expensive. Therefore, CPVC primarily gets selected where such benefits are required, such as in certain chemical or relatively hot liquid services.

CPVC shares most of the features and properties of PVC; it, too, is readily workable, including by machin-ing, welding and forming. However, CPVC requires spe-cialized solvent cement. (For more on such piping, see: “Put CPVC Piping In Its Place,” http://goo.gl/GoIJFT.)

CPVC is more ductile than PVC — allowing greater flexure and crush resistance. Because of its mechanical strength, CPVC is a viable candidate to replace many types of metal pipes in conditions where susceptibility to corrosion limits metal’s use.

Polypropylene (PP) is rugged and unusually resistant to many chemical solvents, bases and acids. It is one of the lightest plastics used in piping systems and comes in various forms. For example, fiber-reinforced-polymer (FRP) wrapped piping combines the excellent chemical resistance of PP with the mechanical strength of FRP.

STRAINS AND STReSSeS

Thermoplastic piping requires flexibility analysis that incorporates appropriate elastic behavior. In many systems, the strains generally will produce stresses of the over-strained (plastic) type, even at relatively low values of total displacement strain. Often, the displacement strains (those due to thermal movements) will not cause immediate fail-ure but may result in detrimental distortion. Progressive deformation may occur upon repeated thermal cycling or with prolonged exposure to elevated temperatures.

Piping layout often offers adequate inherent flexibility through changes in direction, wherein displacements chiefly produce bending and torsional strains of low magnitude. The amount of tension or compression strain (which can produce larger reactions) usually is small.

Where piping lacks inherent flexibility or is unbal-anced, you must provide additional flexibility by one or more of the following means: bends, loops or offsets; swivels or similar; and other special devices and arrange-

ments such as flexible joints. Choose corrugated, bellows or slip-joint expansion joints only where other solutions aren’t feasible.

While different codes and specifications give some general guidance to ensure adequate flexibility, they often don’t provide either specific stress-limiting criteria or particular methods for stress analysis of non-metallic piping systems. This is because of the significant differ-ence of the stress-strain behavior of non-metals versus metals. In particular, Poisson’s ratio varies greatly for the various plastic materials and temperatures; the simplified formulae used as the piping code design basis for stress analysis of metallic piping may not be applicable or valid for some non-metals. Certain codes require the piping system layout to provide substantial flexibility to ensure minimizing displacement stresses; while this approach should allow for a high degree of safety, it isn’t always cost effective. The reality is stress and flexibility analysis of non-metallic piping often depends more on the engi-neer’s experiences and knowledge of specific non-metallic piping under study.

Key CoNSIDeRATIoNS

Temperature is an important parameter. Each thermo-plastic generally has a fixed maximum service tempera-ture, which identifies the upper limit to which pipe may be heated without damage. When heated above this tem-perature, the pipe will soften and deform. Upon cooling, it will harden to the deformed shape and dimensions.

Another important factor for plastic piping is the long-term hydrostatic strength. This — which serves the basis for the piping’s design pressure — is determined by finding the estimated circumferential stress that, when applied continuously, will produce failure of the pipe af-ter around 100,000 hours (say, about 11 years of continu-ous operation) at a specified temperature. In addition, design calculations generally include a service factor that takes into account certain variables together with a de-gree of safety appropriate to the installation. The service factor most often reflects long system life (say, about 40 or 50 years). This design method usually doesn’t include

9

the fittings, joints or cyclic effects such as liquid hammer. Most pressure ratings for thermoplastic pipes are cal-culated assuming a water environment; so, adjustments usually are needed for other fluids. As the temperature rises, the pipe becomes more ductile and loses strength; therefore, you should decrease the rating to allow for safe operation. These factors differ for each pipe material.

Aging can degrade the physical and chemical proper-ties of plastic piping, and generally depends upon tem-perature. The changes can occur naturally through normal atmospheric or temperature fluctuations and ambient light, or can develop because of conditions in the process, such as elevated temperature of the fluid in the piping. One way to determine the onset of aging is to measure thermal stability (oxygen induction time) using differential scanning calorimetry.

Fire conditions greatly accelerate the degradation of plastics. In the early stages of a fire, most plastics melt and lose their structural shape and strength. As the temperature rises, they chemically decompose, often releasing toxic chemicals. This decomposition happens at a lower tempera-ture than ignition. By the time ignition occurs or is possible, a relatively long period of chemical emission has elapsed. When thermoplastic pipe burns, it releases smoke and toxic gases, provides heat that increases the intensity of a fire, and may offer a path for flame to spread along its length. All organic materials are flammable but this is particularly true of polyolefins. It is well proven that many polymers actu-ally are difficult to ignite; addition of flame retardants can further impede ignition.

Continuous application of load on a plastic material cre-ates an instantaneous initial deformation that then increases at a decreasing rate. This further deformation is called creep. Removing the load at any time leads to an immediate par-tial recovery followed by a gradual creep recovery. However, if the plastic is deformed (strained) to a given value that is maintained, the initial load (stress) created by the deforma-tion slowly decreases at a decreasing rate. This is known as the stress relaxation response. The ratio of the actual values of stress to strain for a specific time under continuous stress-ing or straining commonly is referred to as the effective

creep modulus or effective stress-relaxation modulus. Time significantly affects this modulus. Experience shows that all plastic pipe will creep — with the actual extent influenced by time of loading, temperature and environment. There-fore, standard data-sheet values for mechanical properties may not suffice for some design purposes. The stress-strain responses of plastic reflect its viscoelastic nature. The viscous, or fluid-like, component tends to dampen or slow down the response between stress and strain.

Different piping codes identify special protective con-siderations when using non-metallic piping systems. For instance, some codes recommend safeguards and protec-tion against possible impact because plastic piping systems often are vulnerable to accidental impact or similar dam-aging situations; consider safeguards for any above-ground plastic piping from which a spill or leakage can pose safety or environmental hazards. Take into account the lack of ductility and poor resistance to thermal and mechanical shock of some plastic piping systems and provide proper protection. In addition, during design incorporate meth-ods to minimize the build-up of potentially dangerous electrostatic charges in piping that handles electrically non-conductive fluids.

Plastic fittings present a special problem; the geometry of some fittings can result in complex stress patterns that offer some stress concentrations and amplify the apparent stress cycle. A seemingly harmless pressure cycle thus can produce a damaging stress cycle that eventually can cause fatigue fail-ure. This issue is particularly important in the case of branch fittings such as tees. In addition, the existence of stresses from other sources — for example, bending stresses induced by flexing under hydraulic thrust in improperly supported systems — can aggravate the situation. Because the design of plastic fittings isn’t completely standardized, consult fittings manufacturers for recommended derating factors for cyclic loading conditions. Usually you must consider plastic fittings separately from plastic pipes regarding dynamic loading, cyclic analysis and fatigue.

AMIN ALMASI is a rotating equipment consultant based in Syd-

ney, Australia. E-mail him at amin.almasi@ymail.com.

10

COMPRESSED AIR WITH A VAST PORTFOLIO Powering Your Industry With Extraordinary Solutions Does your blower give you the reliability, energy efficiency, and quality you require – with predictive maintenance options that fit your needs? If not, let’s see how we can work together to make your goals a reality.

www.atlascopco.us – 866-546-3588

Atlas Wastewater Ad 7.875 x 10.5 - CP.indd 1 2/9/16 3:07 PM

11

PRoCeSS PLANTS making pH measurements want accurate, reliable performance with a reasonable elec-trode life expectancy while minimizing maintenance. A quality pH sensor system, when undamaged, cleaned and properly calibrated, will provide such performance.

However, an electrode — even in a process that doesn’t cause coating, plugging, abrasion or any other problems — still requires periodic calibration to cor-rect for sensor aging and non-recoverable changes to the electrode. Because these aging effects happen slowly, you shouldn’t have to calibrate more often than once a month in typical general-purpose applications. The need for more frequent calibration stems from a specific cause, i.e.:

• an aggressive process;• ineffective electrode cleaning;• improper routine calibration;• overly temperature dependent pH; or• incorrect electrode selection.Dirty or faulty electrodes can cause anything from

slow response to a completely erroneous reading. For ex-ample, if a film remains on the pH sensor after cleaning, you might misinterpret the resulting measurement error as a need for re-calibration. Correct cleaning can reverse these changes and, so, is a key maintenance step.

The accuracy of pH measurements depends upon maintenance; maintenance frequency largely depends upon the application. Understanding and addressing the causes of pH measurement difficulties are key to ensuring stable and accurate readings. However, troubleshooting a pH system can pose challenges. The guidelines presented here are a good starting point for tackling problems.

INSIGHTS FRoM CALIBRATIoN

Troubleshooting has four main parameters: 1) asymmetry/zero; 2) slope; 3) measuring electrode impedance; and 4) reference electrode impedance.

Most commercial instruments give asymmetry/zero and slope readings. A pH system with a solution ground/liquid ground and advanced sensor diagnostics also will provide impedance values. Understanding the purpose of each of these values allows you to know where problems lurk — and where to start your search for answers.

The asymmetry potential (AS), also referred to as the millivolt offset, indicates the condition of the reference electrode. Theoretically when the electrodes are placed in a pH-7 buffer solution, the millivolt output from the electrode pair (pH and reference) should be zero. An asymmetry read-ing of 20 means the pH sensor is generating 20 mV instead of the expected 0 mV.

The reference sensor causes most asymmetry problems. Some millivolt offsets stem from potassium chloride (KCl) depletion from the reference electrolyte or poisoning of the ref-erence electrolyte with process solution. When the offset is ±30 mV or more, you should replace the reference electrode. Per the Nernst equation, a change in pH of 1 results in 59.16 mV at 25°C. If your pH system has 30 mV of offset, you are adjusting for an incorrect reading equal to 0.5 pH.

The slope, also referred to as the efficiency of the electrode, is an indication of the condition of the measuring (glass) electrode. The slope is given as a percentage value, with 100% being ideal. A new electrode should have a slope in the up-per-90% range. As the electrode ages and loses efficiency, the slope and response will start to decrease.

The slope value is updated each time you perform a

Succeed at pH TroubleshootingHeed some pointers to properly diagnose the cause of a problemBy Cherlyn Marlow, Yokogawa Corp. of America

12

two-point calibration; you should detect only small changes in the value. (Table 1 lists slope reading issues and actions.) Inadequate cleaning can lead to a coating buildup (Figure 1) that causes a low slope value; to avoid this, clean the electrode as needed with a 5–10%-HCl solution for a minute, rinse thoroughly with clean water, soak and recalibrate. Replace the pH electrode when the slope value is in the mid- to low-80% range.

Readings from the reference impedance (RZ), also referred to as the resistance or the reference junction, can indicate the need to clean out a precipitate blockage forming in the refer-ence junction. (Table 2 lists causes and remedial actions for RZ alarms.) The conductivity of the process solution also influ-ences this resistance.

Typically, a clean reference junction will have a resistance of less than 10–15 kΩ but, in low conductivity solutions, RZ values between 200 kΩ and 500 kΩ aren’t uncommon. When the RZ value starts to approach 30–35 kΩ, the electrode will begin to have a slow upward drift. When the reference imped-

ReFeReNCe IMPeDANCe (RZ) ALARMS

ALARM TYPE REMEDIAL ACTIONS

High

Reference sensor dirty Clean sensors.

Reference sensor depleted Replace or refill sensor.

Damaged/open cable Test cable, replace if bad.

Poisoned Sensor Replace or refill sensor.

Low Cable wet/shorted

Protect electrode cables and connections from moisture. Infiltration can damage the insulation, causing a partial short circuit and erroneous read-ings. This often happens to broken cables. Online instruments with im-pedance control should detect these types of issues.

Table 2. RZ alarms can indicate a precipitate blockage forming in the reference junction.

SLoPe ISSUeS

SlopE SYMpToMS And ACTionS

low

Consumed sensor Sensor is old, check serial number for date and replace if necessary.

Poisoned sensor Try cleaning, soaking in pH buffer and then recalibrating the sensors. If still reading low, replace the measurement sensor.

Dirty sensor Clean, then recalibrate.

Cracked sensor Impedance will be low. Replace the measurement sensor.

Wet or shorted cable Test cable, replace if bad.

High

Improper calibration

For example, calibrating a unit with a US 10 buffer that auto calibrates in NIST 9 buffer will result in a slope above 100%.

Poisoned sensor Try cleaning, soaking in pH buffer and then recalibrating the sensors. If still reading high, replace the measurement sensor.

Static charge Place a ground in the pH buffer and recalibrate.

Table 1. A slope that differs from that expected should spur action.

Figure 1. inadequate cleaning can allow a coating to accumulate, as depicted in these extreme examples, causing a low slope value.

CoATING BUILDUP

13

ance exceeds 100 kΩ, an error message should appear on the instrument connected to the pH sensor.

Some instrumentation also will provide a measuring glass impedance value and corresponding alarms. (Table 3 lists causes and remedial actions for electrode impedance alarms.) These alarms typically serve to indicate faulty equipment.

CoMMoN PRoBLeMS AND ReMeDIeS

Once a system is calibrated and functioning properly, problems still can creep into an application. Here are a few common examples, along with corresponding solutions.

Improper cable preparation. Technicians often cut off excess cable rather than buying the correct length cable — this is not a good idea. To prevent inside radiation or disturbances, a cable comes with a special layer of graphite for screening. Removing this layer is very difficult. Measurement errors and instability commonly turn up in a shortened cable due to a severe loss of isolation resistance between its core and screen. For a glass elec-trode cable, this isolation resistance must exceed 1,000 times the resistance across the glass membrane.

Drifting. Over time, a probe’s reading may drift up or down. As the asymmetry potential (mV offset) becomes greater, the electrode will drift and require more frequent calibration. A common cause is depletion of KCl from the reference electro-lyte. This generally happens when a gel-filled reference elec-trode is used in the wrong process solution, such as high-purity water. When the concentrated reference electrolyte (KCl) and the low conductivity (mineral-free) process solution meet at the reference junction, the process solution leaches salt from the reference electrolyte, causing the reference potential to become unstable (Figure 2).

Another cause of drift is poisoning of the reference elec-trolyte with the process solution. This is most common in ap-plications with process pressure greater than 1 atm. The process pressure overcomes the pressure inside the electrode, forcing process solution into the reference electrode and contaminating the electrolyte. This increases the millivolt offset; replace the reference electrode when the offset value is ±30 mV or more.

Process solutions containing sulfides or sulfur-bearing species can cause drifting by reacting with the silver chloride

eLeCTRoDe IMPeDANCe ALARMS

ALARM TYPE REMEDIAL ACTIONS

HighMeasurement sensor cable disconnected or bad Terminate properly or replace.

Measurement sensor bad Replace measurement sensor.

Low

Cracked measurement sensor Replace measurement sensor.

Wet or shorted cable Replace cable.

Poisoned sensor Clean, soak in pH buffer and recalibrate. Replace if necessary.

Table 3. These alarms often signal that sensor replacement is necessary.

Figure 2. in this type of probe, loss of electrolyte results in expansion of internal bellows as shown in bottom photo.

eLeCTRoLyTe DePLeTIoN

14

in the reference electrolyte, prompting formation of insoluble precipitates in the reference junction. These precipitates result in high electrical junction resistance, which leads to non-repro-ducible diffusion potentials, causing the electrode to have a slow constant upward drift. When the RZ value begins to approach 30–45 kΩ, the electrode will start a slow upward drift.

Noisy readings. Because it is open to the process via the junction, the reference electrode is the weakest link in the mea-surement loop; it is very susceptible to the interference of stray voltages in the liquid. In many applications, the voltage poten-tial of the process liquids is significant and can’t be neglected when making pH measurements. Therefore, you must ground the liquids at the point where the pH value is to be measured, usually by using metal fittings or plastic fittings with a ground-ing (solution ground) electrode of suitable metal.

One way to test if stray voltages are causing errors is to take the pH system offline and place the sensor in a process grab sample. If the sample reading is stable but drifts when placed back online, then stray voltages are an issue. Another test is to install a jumper in the instrument between the reference and the solution-ground terminals. If the reading locks in place, a stray voltage problem exists.

Slow response time. Sluggish performance indicates coating or plugging of the junction, which can be caused by a thin film invisible to the naked eye on the glass sensor.

A variety of cleaning solutions can remove a coating. However, a 5–10% solution of HCl usually works well. Follow these steps:

1. Rinse the sensor in plain water to remove any heavy process coating.

2. Immerse the electrodes in the cleaning solution for one to two minutes, agitating them regularly.

3. Use a soft brush to clean off coating deposits, taking care not to damage the electrodes.

4. Rinse the electrodes thoroughly with clean water to avoid contamination of the calibration buffers.

5. Soak for two to three minutes in water.As the electrode ages and loses efficiency, response time

slows and the slope value decreases. As already noted, a low slope value indicates improper cleaning before calibration.

The pH measurement is wrong online but correct in buffers. This common problem is called diffusion potential. If the sensor junction is plugged, the electrical contact between elec-trolyte and process is poor, so diffusion potential is measured directly as an error. The chemical composition of a pH buffer

differs from that of the process liquid; therefore when the junc-tion is in bad condition, this error is calibrated in the pH buffer solution but varies for the process solution. An easy check is to look at the diagnostic information in the pH instrument — high asymmetry potential or low slope indicates this problem. Ground loop current caused by a pH sensor without proper solution grounding also can be the culprit.

Cracked membranes. A minute invisible crack in the mem-brane of a glass electrode, often caused by frequent temperature shocks, can cause measurement errors. For example, if an instrument reads 0 mV it will show pH 7, but if the sensor is placed in a buffer with pH 4, the instrument still will read pH 7. For neutralization processes where the set point typically is pH 7, this is very critical and a dangerous situation. Without additional diagnostic checks, the error will remain undetected. Those pH systems with a solution ground and online imped-ance-sensor diagnostics frequently monitor the impedance of the pH membrane via the solution ground, and will generate an alarm in the case of a broken glass membrane (Figure 3).

Dry electrodes. When a pH electrode dries out in storage or in a process, its performance suffers. The pH reading may become slow and erratic and can shift upscale, resulting in a shortened span. A dry pH electrode possibly can have its gel layer rejuvenated. Short exposures to dry conditions caused by calibration, troubleshooting routines or a batch refill of a process tank usually require no extra handling to regain a fully functioning electrode.

You can treat longer exposures, such as being uncov-ered in an empty tank or sitting on a bench for 12 hours or more, in several ways. The first is to soak the electrode in a 4.0-pH or 7.0-pH buffer solution, a reference fill solution or tap water. The length of soaking depends upon how long the electrode was left dry; soaking typically lasts 30 minutes for short dry periods and to 24 hours for more severe instances. Alternatively, you can place the electrode into service immediately with the understanding that some

Figure 3. Minute cracks in glass lead to measurement errors and often remain undetected without additional diagnostic checks.

CRACKeD MeMBRANe

15

measurement uncertainly will exist until the electrode recovers. During this recovery period, a pH calibra-tion will be marginally helpful. Because the gel layer is changing with exposure to process fluids, to achieve best accuracy you must perform a second calibration within 8 to 24 hours. Typically, heating dried-out sensors to 80°C in electrolyte and then allowing them to cool to room temperature will restore the sensors.

Laboratory and process measurements don’t match. When this happens, the normal inclination is to consider the inline measurement wrong. You can determine the actual fault by following these steps:

• An accuracy of ±0.1 pH for each instrument normally is acceptable, with further troubleshooting only re-quired when the total uncertainly exceeds ± 0.2 pH.

• Are both instruments accurate? Validate each by measuring two or three fresh buffer solutions. Don’t make any adjustment. Record the values and judge the results. If one is wrong by 0.1 pH or more, calibrate it and repeat the test.

• Compare apples with apples by ensuring the inline and the laboratory instruments are measuring the same sample at the same pressure and temperature. Don’t measure in the lab at reference temperature when the inline measurement is at process temperature. Compensate the inline instrument temperature by taking a hot sample, inserting the sensor and letting it cool to 25°C. The reading should become stable if there’s proper temperature compensation. If the read-ing changes, calculate the Δ pH/Δ temperature, and program this coefficient into the instrument if that’s possible.

• Consider the process properties. For example, with boiler feedwater, the process solution will be ultrapure water with traces of ammonia or morpholine that increase the pH. So as soon as the sensor is exposed to ambient air, the pH will drop due to absorption of atmospheric carbon dioxide.

START WITH THe ReFeReNCe eLeCTRoDe

When troubleshooting, remember that the reference electrode accounts for 80% of all typical pH-measurement problems.

As discussed, the main issues are:• Process solution enters through the dia-

phragm or junction and poisons the Ag/AgCl element or electrolyte.

• The diaphragm/junction becomes coated or plugged.• The internal electrolyte (KCl solution) becomes de-

pleted.These all undermine proper operation of the reference

electrode, causing instability and inaccuracy. You can solve most problems through basic maintenance:• Coating of the glass membrane (sluggish readings):

Clean the membrane.• Coating or plugging of reference junction (open circuit):

Perform cleaning or use polytetrafluoroethylene junc-tions so particles don’t adhere. Positive pressure and a flowing reference will help prevent coating and plug-ging.

• High resistance between electrodes (pure water or an open circuit): Ensure the flowing reference provides enough KCl ions to achieve a conductive path between the reference and glass electrode to complete the mea-suring circuit.

• Pressure spikes drive process solution into the reference electrode, forcing out electrolyte and causing a change in KCl concentration and a loss of a stable reference voltage: One way to combat this is to use a reference system having a flowing junction type that maintains a constant flow of electrolyte out of the sensor at all times.

• Some ions (e.g., sulfides, cyanides and bromides) react with Ag+ to form silver sulfide, which deposits in or on the junction, plugging it and attacking the internal reference pin: A double-junction-style reference can delay this poisoning or a positive-pressure version of the sensor can prevent the poisonous ions from migrating into the reference.

Ultimately, poisoning of the Ag/AgCl element, plug-ging of the junction, and depletion of internal KCl solution all are key measurement concerns that you must address. You can reduce the effects of these problems by increas-ing maintenance, including the frequency of cleaning and calibrating the electrodes, as dictated by the aggressiveness of the process.

CHeRLyN MARLoW is a product manager for Yokogawa Corp. of

America, newnan, Ga. E-mail her at cherlyn.marlow@us.yokogawa.com.

ReLATeD CoNTeNT oN CHeMICALPRoCeSSING.CoM“Prevent pH-related Problems,” http://goo.gl/X4wrcA “Measure pH Accurately,” http://goo.gl/5fxe1c “Forestall pHrustrations,” http://goo.gl/3EHKUK “Maintain Accurate pH Readings,” http://goo.gl/3A4xRd “Take Some Basic Steps with pH Measurements,”

http://goo.gl/wCWMCi

16

Proven Water Solutions. Delivered.

Complete Membrane Bioreactor Systems

Smith & Loveless refines the Chemical Processing Industry’s water use while preserving its capital and operational reserves with our proprietary MBR, fixed-film and extended aeration wastewater treatment systems. We design and build packaged, containerized/mobile and field-erected systems that cost-effectively treat wastewater and stormwater — including for water reuse. Our compact packaged and erected designs intentionally preserve facility space while limiting capital and long-term O&M costs.

Request a budget and design proposal today!

CALL 800.898.9122 VISIT SmithandLoveless.com

Complete Wastewater Treatment & Container System Solutions

Large-Flow, Field-Erected Systems

Proven Fixed-Film Systems

17

WATeR IS becoming an increasingly precious resource, in some areas more than others. As such, industries that rely on water are interested in its conservation. Water can be conserved in a variety of ways, but one efficient method is to operate closed-loop systems such as those used to reclaim steam condensate for reuse as boiler feed water.

Within the petrochemicals and power generation indus-tries, this reclamation is a common practice. Such processes can improve plant economics by offering a ready source of already heated water, often >80°C. Because it is a conden-sate, it typically does not require further demineralization by ion exchange. Recycling the condensate also eliminates environmental issues associated with disposing of contami-nated water.

However, when recycled condensates are to be used as boiler feed water for medium- or high-pressure boilers for process or power generation applications, a downstream ion exchange unit must be installed to prevent trace quantities of minerals leached from the activated carbon from reduc-ing boiler and turbine efficiency.

Condensate feed waters frequently contain hydrocarbon contaminants that are entrained with the water from pro-cess equipment, for example, from leaking seals on rotating equipment such as pumps or from chance contamination by associated hydrocarbon process streams. This hydrocarbon contamination must be removed from boiler feed waters to avoid coking of these contaminants during heating in the boiler, which produces carbon deposition within the boiler tubes. This deposition will reduce heat transfer and overall boiler efficiency. Hydrocarbon concentrations may be in excess of 6 mgL-1 but more typically are 2–4 mgL-1.

In flash boilers and welded water tube boilers, coking occurs on the surfaces where heat flow is highest, causing tubes to overheat and distort or burst. For boilers using expanded joints, the problem may be worse as the oil pen-etrates the joint quickly and causes the boiler to leak. Sand blasting can take care of boiler decoking.

However, refinery process boilers frequently need more extensive decoking operations. This may require the plant being taken off-line and boilers burned out with high-tem-perature oxygen or air. Keeping the boiler clean of hydro-carbons can be a major maintenance task.

Condensate typically contains <5 mgL-1 of hydro-carbons, and at these concentrations active carbon is an efficient adsorption medium for entrained hydrocarbons. It is used widely to treat condensate because of its high removal efficiencies, frequently greater than 95% removal, and some carbon adsorbents have hydrocarbon uptakes of ~30wt % to saturation.

At concentrations greater than ~ 6 mgL-1, carbon retains its effectiveness, but additional process equipment may be required. This equipment may include coalescers or pre-filter beds to prevent premature fouling of the carbon adsorbent by oil droplets.

CARBoN ADSoRBeNT PRoCeSS

To understand how the carbon adsorbent process works, take the example of a generic refinery-scale condensate treatment process. This type of system uses twin adsorbers that remove 2–4 mgL-1 (saturation level) of dissolved oil from hot condensate at 45°C upstream of an ion exchange resin (IER) treatment bed feeding an 85°C upstream heat

Treat Condensate Water with Activated CarbonHigh-purity coconut-based carbons reduce silica leach, rinsing requirements

By Steven Ragan, Jacobi Carbons Group

18

exchanger/boiler return loop. For the process parameters, see Table 1.

Start-Up. After the activated carbon has been put into the adsorber, initial washing is undertaken with hot conden-sate in downflow at a rate of 2 m3 condensate/m3 activated carbon per hour for a period of 24 hours. The initial wash-ing’s effluent is discarded and drained.

Following this initial washing stage, the filter is back-washed to classify the bed. The sequence of washing is summarized below:

• 30 min. of backwashing at a bed expansion or 30%• 20 min. of backwashing at a bed expansion or 20%• 5 min. of backwashing at a bed expansion or 15%• 5 min. of backwashing at a bed expansion or 10%Operation. After completing the backwashing cycle,

filtration in the downflow direction begins. The optimal process flow rate should be controlled to maintain design contact time, and the effluent water’s silicon content should be monitored.

If the effluent water’s initial silica measurements do not comply with the plant’s operational requirements, additional washing with hot condensate, in downflow, at a rate of 2 m3 condensate per m3 activated carbon per hour for a further period of 48 hours (96 bed volumes of hot condensate) must be undertaken.

Effluent from the second washing then should be dis-carded to drain. Backwashing may be repeated as detailed previously and as deemed necessary.

Should filter head loss become excessive, adsorbers may be backwashed to 30% expansion as in above.

Operational Issues. Principal operational issues typi-cally concern hydrocarbon loads and silica leach from the activated carbon.

As mentioned previously, should hydrocarbon loads exceed 5 mgL-1, the process stream should be pre-filtered to prevent premature fouling of the carbon adsorbent by liquid oil droplets. This filtration should use anthracite or similar carbonaceous material as the pre-filter media. Sand should not be used as a filtration medium as it will markedly in-crease the condensate’s silica content.

Remember that medium- to high-pressure boilers using recycled condensate as boiler feed water require a down-stream ion exchange unit. The plant operator is responsible for ensuring that the amount of silica leaching from the activated carbon is within the required limits before starting the boiler/turbine system.

Activated carbon contains silicon in its ash, so the plant operator must wash it out during the start-up procedure to avoid silica leach that is detrimental to high-pressure boilers and power generating turbine systems. Silica deposits in boiler tubes creates heat exchanger problems similar to cok-ing from oil in condensate.

Silica deposits on turbine blades in power generation equipment is more problematic as it can unbalance the turbine’s high-speed rotation and induce catastrophic axle failure.

It is recommended to use specially prepared activated carbon that can withstand multiple washings to help combat this problem. It removes soluble silica ash and is characterized by low silica release into condensate after initial backwashing.

ACTIve CARBoNS FoR HyDRoCARBoN ReMovAL

One type of adsorbent is a washed, high-purity, high-activ-ity granular activated carbon produced by steam activation from selected coconut shell charcoal. This adsorbent is

PRINCIPAL PRoCeSS PARAMeTeRS

Volume flow rate: 50 m3 hour-1 (total)

Carbon filters: 2×10.5 m3 adsorbers in sereis (21 m3 total)

Contact time: 2×12 min. (24 min. total)

Concentration oil:

Inlet: >2.0 mgL-1

Outlet: <0.18 mgL-1

Si02 ex GAC: <0.05 mgL-1

Si02 ex IER: <0.03 mgL-1 (mean value)

Table 1. Above are the paramaters for system using twin adsorbers that remove 2–4 mgl-1 of dissolved oil from hot condensate at 45°C upstream of an ion exchange resin treatment bed feeding an 85°C upstream heat exchanger/boiler return loop.

19

particularly suited for ultra-pure water treatment systems requiring water of low conductivity and exceptionally high purity. This material also is specifically recommended to remove heavy hydrocarbons from condensates. The washing process reduces soluble silica (SiO2) from the activated car-bon’s matrix carbon to prevent leaching into the condensate. (Table 2 highlights the typical properties of coconut shell adsorbent.)

CoMPARISoN oF ACTIvATeD CARBoNS

Table 3 compares properties of high-purity activated carbon from coconut shell charcoal, 0.8-mm peat pellet, and acid-washed 0.8-mm coal-based pellet. The comparison is based on commercially available data.

The comparison table shows that coconut-based carbon has a 20% greater density than coal- and peat-based pellet-ized products. Because of this, an adsorber loaded with a coconut carbon will exhibit a greater hydrocarbon capacity overall as it also contains 20% more carbon. The product’s high hardness results in less carbon attrition in the adsorber and lower release of fine carbon particles into the conden-sate during adsorber operation.

However, it’s important to note that in condensate treatment applications, using a high-purity coconut carbon product (75% less ash than the acid-washed coal and peat pellets) will result in less silica leach into condensate, reduc-ing rinsing requirements as well as lowering the risk of potential boiler or turbine problems. Coal- and peat-based carbons produce greater silica leach than coconut-based car-bon products as their ash composition has a large alumino-silicate content and even after acid washing can be reduced to only ~3% residual ash.

STeveN RAGAN is research and development director for Jacobi

Carbons Group. He can be reached at steven.ragan@jacobi.net.

CoCoNUT SHeLL ADSoRBeNT PRoPeRTIeS

Iodine number mg g-1 1,050

BET surface area m2 g-1 1,100

Apparent density kg m-3 510

Bed density, backwashed and drained

kg m-3 435

Total ash content wt % 0.5

Si leach after 10 bv mgL-1 0.1

Moisture content - as packed wt % 4.0

Acid soluble matter wt % 0.01

pH 6

Ball-pan hardness number % 99

Table 2. Coconut shell adsorbent properties typically feature a large adsorptive capacity for oil and levels of silica release into condensate that are much lower than other carbons.

CoMPARe ACTIvATeD CARBoNS

Typical Property Coconut Shell Charcoal 0.8-mm Peat PelletAcid-Washed

0.8-mm Coal Pellet

Backwashed density kgm-3 435 380 380

Ash wt % 0.5 8 5

Iodine no. mg g-1 1,050 1,050 1,050

Surface area m2g-1 1,100 850 900

Hardness % 99 94 92

Moisture wt % 4 2 2

Table 3. This comparison shows coconut-based carbon has a 20% greater density than coal- and peat-based pelletized products.

20

eHandbooks Check out our vast library of past eHandbooks that offer a wealth of information on a single topic, aimed at providing best practices, key trends, developments and successful applications to help make your facilities as efficient, safe, environmentally friendly and economically competitive as possible.

Upcoming and On Demand Webinars Tap into expert knowledge. Chemical Processing editors and industry ex-perts delve into hot topics challenging the chemical processing industry today while providing insights and practical guidance. Each of these free webinars feature a live Q&A session and lasts 60 minutes.

White Papers Check out our library of white papers covering myriad topics and offer-ing valuable insight into products and solutions important to chemical processing professionals. From automation to fluid handling, separa-tions technologies and utilities, this white paper library has it all.

NEW!! Minute Clinic Chemical Processing’s Minute Clinic podcast series is designed to tackle one critical issue at a time — giving you hard-hitting information in just minutes.

Ask the Experts Have a question on a technical issue that needs to be addressed? Visit our Ask the Experts forum. Covering topics from combustion to steam systems, our roster of leading subject matter experts, as well as other forum members, can help you tackle plant issues.

Visit the lighter side, featuring drawings by award-winning cartoonist Jerry King. Click on an image and you will arrive at a page with the winning caption and all submissions for that particular cartoon.

TOP COmiCal PrOCessing

ADDITIoNAL ReSoURCeS JOiN Us ON

sOCiAl MEDiA!