INTERPRETATION OF DEPOSIT WEIGHT DENSITY ANALYTICAL RESULTS AS A

MEASURE OF BOILER TUBE CLEANLINESS

George Totura Tom Spry

Ondeo-Nalco Company Smurfit-Stone Container Corporation

ABSTRACT

Boiler tube overheat failures can often be avoided via a testing program whereby the accumulation of
waterside deposition is quantified via an analytical procedure known as Deposit Weight Density (DWD)
testing. But there are three widely recognized DWD test procedures in use today, and there is enough
variation in the results that different decisions might result, depending on which test procedure is used.
The three current analytical procedures in use are:

Mechanical Removal: A vibrating tool is used to physically remove the waterside deposition from a
measured area of tubing. The scraped deposit is collected and weighed.

Acid Etching: An acid is used to dissolve the deposition from a measured area of tubing, with the final tube
weight being compared to the initial.

Glass Bead Blasting: Non abrasive glass beads are used to dislodge deposition from tube surfaces, with the
tube being weighed before and after removal to develop the correlation.

A frequently used evaluation of DWD testing results has been developed by Babcock and Wilcox
Company, a foremost US boiler manufacturer, and used for many years in the industry. Depending on the
boiler’s operating pressure, a DWD result may or may not suggest the need for deposit removal. Certain
critical boilers, such as paper mill kraft process recovery boilers may be judged more conservatively, with
lower DWD results seen as requiring deposit removal.

The problem as it currently exists is that the different testing procedures may result in different conclusions
being reached as the whether cleaning is required, based on the fact that different test procedures yield
different results. The purpose of this paper is to discuss the three procedures in current use, and suggest a
way of interpreting the results so that accurate operating decisions about the need to chemically clean the
boiler can be made.

INTRODUCTION

The accumulation of deposits on the waterside of boiler tubes is a major concern for paper mill operating
teams. It is well known that deposits can be insulative, causing the tube metal temperatures to rise above
the recognized stability threshold value and the tubes can then fail. Less obvious, but no less troublesome
is the fact that waterside deposits can allow for localized chemistry conditions to develop and set the stage
for corrosion damage.

There are a number of ways in which the accumulation of waterside deposits is evaluated. Visual
inspections tell us something about waterside cleanliness, but the areas of the boiler that can be readily seen
are often the least important in terms of heat transfer. Inspection capabilities can be enhanced by electronic
means such as TV cameras and borescopes, but again it is often difficult to see the areas in the boiler of
greatest interest and concern. A more drastic step is the removal of a tube section for laboratory
determination of the deposit build-up. This destructive test method is not popular with operating crews

Presented at TAPPI Fall Technical Conference – San Diego, CA, September 2002.
because of the need to weld in replacement tubing, but it may provide the most quantifiable means of
assessing boiler waterside conditions. It is also a challenge to decide from where tube sections should be
extracted, but boiler manufacturers offer guidance in this process by locating the areas where heat transfer
is the highest and deposits are thus the most likely and most damaging.

But even after collecting a tube specimen from a boiler the powerhouse management team is faced with yet
another concern. What analytical test method should be used to evaluate the accumulation of deposition in
their tube(s), and what do the results mean in terms of possibly needing to chemically clean the boiler,
either now or at some time in the future? It is the purpose of this paper to give guidance to the operating
crews of paper mill boiler systems and those who assist them in answering these complex and crucial
questions.

UNITS OF MEASURE

The quantification of accumulated deposits on a boiler tube is often called the determination of the Deposit
Weight Density, or DWD. The unit of measure used in North America to express DWD results is grams of
deposit per square foot of tube surface area. This is clearly an inconsistent term in that it mixes a metric
unit (grams of weight) with an English unit (square feet of area), but it is the one that is in common use.
The purely metric equivalent of this is milligrams of deposit per square centimeter of deposit. It is a
fortunate occurrence that in converting between these expressions that there is less than an 8% difference,
so they are used interchangeably in many cases. For the purposes of this paper, we will express the results
in terms of grams per square foot, realizing that these can be precisely converted to milligrams per square
centimeter by multiplying by 1.076.

TEST METHODS

There are three well-documented test methods for evaluating the DWD accumulations on tube surfaces.
All have been rigorously evaluated, and will provide consistent results in the hands of a well-trained
analyst. Unfortunately, the different natures of the three test methods make it inevitable that the results will
vary between methods, possibly to the point where a different operating decision might be reached solely
on the test method employed. These differences should become apparent as each of the three test methods
is described in detail.

For all three test methods it is a common and useful practice to split the tube lengthwise and evaluate the
hot and cold sides separately. It is common for there to be more deposition accumulated on the hot side of
the tube, since it is exposed to more heat transfer from the fireside. The difference in heat input between
the hot and cold sides is typically not so severe as to induce any steam/water phase separation concerns.

Mechanical Method

A mechanical technique for removing deposits from a measured area of tube surface is described as ASTM
D-3483, Method A. The essence of this procedure is that a two-step process for deposit removal is used to
clean the tube down to the protective magnetite layer. The first step of the process is to define a test area
from which the deposit is to be removed. The analyst then applies a scraping procedure, utilizing a
screwdriver, knife, or laboratory spatula to remove the more loosely adherent part of the deposition present
from the tube surface. This fraction of the deposit accumulation is sometimes reported as the brushable
fraction, characterizing it as being less adherent to the tube surface. The second step of the process is to
use essentially a jeweler’s engraving tool to scratch or scrape the more adherent part of the deposit from the
measured test area. This second part of the deposit fraction may be called the scrapeable fraction. The
total deposit accumulation, what really matters in terms of the possible need to clean the boiler, is the sum

Presented at TAPPI Fall Technical Conference – San Diego, CA, September 2002.
of the two fractions. Once the mass of deposit is determined is an excellent practice to send the
accumulated deposit to another part of the laboratory for elemental analysis.

While we often consider the DWD number to be of paramount importance, differing deposit compositions
may transfer heat very differently. An identical DWD number for a silica-bearing deposit may be much
more insulative and therefore troublesome than a mixed metal oxide (iron plus copper) matrix. The ability
to learn the chemical composition of the deposit removed is a uniquely beneficial aspect of the Mechanical
Method.

A number of issues contribute to induce variability into the results of this mechanical procedure. First of
all the analyst needs to select the area, almost always something less than the entire tube sample collected,
over which to do the testing. Then the analyst can apply more or less force on the tool during the removal
of the less adherent deposit fraction. Also, the analyst can apply more or less force in support of the
engraving tool, possibly to the point where tube metal, not deposit, is removed and measured as deposit
weight. The vibrating nature of the tool also generates dust and debris that may escape the analyst’s efforts
to collect all removed materials for mass measurement, lowering the resulting DWD.

In general terms, the Mechanical ASTM D-3483, Method A technique produces the lowest apparent DWD
results, and the results of this test procedure also tend to show the most variation from test to test.

Chemical Method

The companion chemical approach to DWD measurement is the ASTM D-3483, Method B procedure,
which is often called the Atwood–Hale protocol. Following a defined preparation of the tube’s fireside
surface to eliminate any loss of tube mass not associated with waterside deposition, the tube section is
initially weighed. This method then requires the tube to be placed in a heated inhibited hydrochloric acid
bath of specified strength and temperature. The waterside deposits are dissolved into the acid solution,
although some analysts reportedly give the process a mechanical assist by doing a degree of deposit
scraping after an initial soaking period. The intent of the mechanical assist part of this test is to limit the
time exposure in the acid bath, during which base tube metal, not deposit, may be dissolved, and thus
contribute to an erroneously high test result. The test protocol does not specify a time exposure period,
rather it is left to the analyst to determine a point at which the tube is “clean”. After rinsing and drying, the
tube specimen is re-weighed to ascertain the mass of deposit theoretically removed. Careful measurement
of the tube specimen’s dimensions allow for the development of the test area, and then the mass per unit
area value can be calculated.

A number of factors contribute variability in the results derived from the Chemical Method, most of them
increasing the apparent DWD value so that this method always reports the highest apparent deposit
accumulations. Hydrochloric acid is an indiscriminate dissolver of metals, and even though a film-forming
inhibitor is added to reduce the amount of base metal wastage, some must inevitably occur. Any sharp
metal edges or burred edges resulting from the tube sample preparation are more easily dissolved by the
acid, and contribute to the apparent mass removed. Those of us with a chemical cleaning background know
it is important to control the iron concentration in the acid solution, otherwise a degree of ferric ion
corrosion of the base tube metal will also dissolve more metal that is not actually a deposit. And lastly, the
acid must also dissolve the protective magnetite film we intentionally develop on the tube surface and
report this as deposit loading, even though we really want and need it on the tube surface.

So almost without exception, tube samples evaluated via the Chemical Method will suggest the presence of
a higher deposit accumulation, suggesting the need for more frequent acid cleanings to be done to maintain
waterside cleanliness. This procedure is also not a favorite of laboratory managers, who need to deal with
the volumes of metal-laden acidic waste generated. It is not possible to ascertain the deposit’s composition
after removal in the acid bath, so for the important step of determining the chemical composition another
part of the submitted specimen must be scraped to some degree similar to the Mechanical Method. This
material can then be submitted for analysis.

Presented at TAPPI Fall Technical Conference – San Diego, CA, September 2002.
Glass Bead Blast Method

While not having the rigorous sanctioning of the ASTM, the Glass Bead Blast method is coming into more
widespread use for the development of DWD results. Sample preparation for this method is similar to the
Chemical Method, in that the fireside surface is cleaned so as to not cause additional weight loss, and the
specimen is pre-weighed. Somewhat similar in concept to sandblasting, a gritty material is entrained into
an air stream and contacted against the previously prepared tube, waterside and fireside. It was learned
during the development of this protocol that glass beads would not be so aggressive as to remove base tube
metal and induce false high readings. As in the chemical method no precise exposure time is specified, but
exposure times of 15 – 20 seconds are the norm. The contact is typically conducted in a small cabinet or
booth, so that dust and grit are not spewed into the laboratory air space.

Experience in the ten-plus years that this procedure has been in use is largely positive. The numerical
results derived are most always in between those of the Mechanical and Chemical Methods. There is little
analyst involvement or decision making in conducting the test, only the occasional extra exposure time
required when a particular deposit is especially tenacious, so the results tend to be the most consistent. The
glass beads in widespread use can be shown not to be erosive to base tube metal, so false-positive results
are minimized. And larger test specimens are more readily measured reducing the error involved in
measuring small test areas.

As a result of these factors, the Glass Bead Blast Technique is coming into wider and wider usage, and
could be considered as the standard technique for making decisions regarding the need for boiler chemical
cleaning.

RESULTS INTERPRETATION

One of the important reasons a deposit weight density test may be conducted is to provide input to a
decision on the necessity of chemically cleaning the boiler. There are nearly as many strategies on
determining when a boiler needs to be cleaned as there are operating plants. Many paper manufacturing
corporations have established policies based only on the elapsed time since the previous cleaning or other
operating indications that suggest that the boiler needs to be cleaned. Deposit weight density is
recommended one of the factors that should indicate the need for a cleaning, but certainly not the only one.

The observation of an acceptable deposit weight density test result is certainly a positive, but cannot be
used as the complete justification not to clean a boiler. While tube specimens are often collected from
areas in the boiler where the maximum deposit accumulation is expected to occur, boilers have been seen to
have significant deposit build-ups in unexpected areas. A recent trend observed at a number of kraft mills
is for there to be heavy deposits forming in the area immediately above the field weld line where the
composite tubing ends and mild steel tubing begins in their recovery boilers. This is an area where the heat
input, a factor known to enhance deposit build-up, is lower. The fact that this phenomenon has occurred in
multiple mills makes it unlikely that it is a single mill’s operating conditions that are causing the
deposition. So while it must be accepted that a single DWD number cannot be interpreted as providing a
clean “bill of health” for a boiler, at some point higher values must also be understood to provide a
conclusive justification for boiler chemical cleaning.

One major international paper manufacturing corporation has adopted the following table as it guideline for
using deposit weight density data in the decision to chemically clean a boiler. It is worth noting that a
number of factors are involved in the decision process:

Presented at TAPPI Fall Technical Conference – San Diego, CA, September 2002.
DWD Technique. The two rows of the data table show that this corporation accepts the conclusions
developed earlier in this paper that different deposit weight density measurement techniques will likely
produce different results.

Critical Nature of Recovery Boilers. Kraft process recovery boilers are set aside as a special grouping
where less deposition can be tolerated because of the absolute necessity of avoiding tube failures in these
units.

Boiler Operating Pressure. Higher pressure boilers maintain metal temperatures closer to the 850oF
threshold at which mild steel begins to deteriorate. For this reason boilers operating a in the 600 – 1300 psi
operating range are determined to be less tolerant of deposit accumulation than lower pressure units.

Table I. Boiler Deposit Accumulation Threshold Requiring Acid Cleaning

Boiler Acid Cleaning Threshold Value [grams per square foot]

DWD Technique All Recovery Boilers Power Boilers Operating 600 – Power Boilers <
1300 psi 600 psi
Glass Bead Blasting or 20 40 80
Mechanical Scraping
Followed by Glass Bead
Blasting
Acid Removal or 25 50 90
Mechanical Scraping
Followed by Acid Removal

ACID CLEANING DECISION PROCESS

The recommended process for determining the necessity of chemically cleaning a specific boiler is what
could be called the “Body Of Knowledge” Method. This approach attempts to not rely on a single data
point or consideration, but blends a series of findings into a sound engineering decision. There are five
components to the Body of Knowledge Method of determining the need to chemically clean a boiler. They
are:

Time since most recent cleaning. Since almost every boiler feedwater contains potentially depositable
materials, the longer the run since the previous acid cleaning the more deposits can reasonably be expected.
Imposing too short a time between acid cleanings ignores the benefits of many operational and chemical
treatment strategies, and exposes the boiler to more aggressive solvents which could potentially result in
tube thinning and eventually require boiler retubing. But some recently developed data shows that even if
the total deposit mass may not change significantly over time, the nature of the deposits may, which can
make the next acid cleaning more difficult and time consuming.

Table II is the data from a series of deposit weight density examinations on tube sections from a Pacific
Northwest US paper mill:

Table II: Changes in Boiler Deposit Accumulations over Time

DWD Method Deposit Weight Density Data [grams per square foot]
2001 2001 1997 1997 1994 1994
Hot Side Cold Side Hot Side Cold Side Hot Side Cold Side
Mechanical 2 3 6 3.5 5 7.5
Removal
Acid Etch 8 10 4 5 3.5 2.5
Total Deposit DWD 10 13 10 8.5 8.5 10.0

Presented at TAPPI Fall Technical Conference – San Diego, CA, September 2002.
Several important conclusions can be drawn from this data:

1. The total deposit accumulation has not increased significantly in the period since 1994. There would
be no compelling need to chemically clean a 600 psi recovery boiler based on this data.

2. The apparent nature of the deposit may be changing. While in 1994 the majority of the deposition was
readily removed by the mechanical removal method, the fraction of the deposit removed by this
technique has dropped significantly, from 58 – 75% removed in 1994 to 20 – 23% removed in 2001.
This suggests that the deposition may be getting more tenacious due to the extended time since the
previous cleaning.

In the chemical cleaning business it is known that the mechanism by which boiler deposits are removed is
not just dissolution of the adherent matrix. In many cases the solvent preferentially attacks the interface
between the accumulated deposits and the tube surface. This concept is supported by the frequent
observation of loose chunks of deposit in lower drums and headers following an acid cleaning procedure.
In boilers that have gone many (15+) years since the previous acid cleaning there have been a number of
cases of difficult cleanings reported when the procedure was finally performed. The point could be made
that the long-term presence of deposits in boilers can result in more adherent deposits that are most difficult
to remove. This makes a clear case that the time component of the cleaning Body of Knowledge decision
path is important.

Deposit Weight Density. While the amount of deposits built up on tube surfaces varies greatly within a
boiler, the observation of an excessive DWD test result is a clear indication that a chemical cleaning is
warranted. Some provisions should also be made for the deposit composition to account for the varying
heat conducting capability of different deposit compositions.

Unfortunately the converse statement is not necessarily true, that the determination of an acceptable DWD
result means that cleaning is not required. It may be that the tube specimen was not collected from where
the maximum deposition is accumulating. The boiler manufacturers can provide guidance on the areas
where the greatest heat transfer is occurring in the boiler, but as was reported earlier, there are cases where
deposition is being seen in unexpected locations. A major part of the earlier part of this paper has been the
consideration of DWD data leading to a decision on the need to clean the boiler. It is understood that this is
a destructive test method that requires replacement tube sections to be installed in the boiler, but the data
that can be developed is often the best quantitative determination of boiler cleanliness.

Visual Inspection. Annual or more frequent internal boiler inspections are a staple of the industry. Many
conclusions about boiler cleanliness have been developed on the basis of drum inspections. But the drums
are often the areas where the minimum deposition can be expected, not the maximum or worst case
scenario we need to consider. The heat flux into boiler drums may be lowest of any area in the unit.
Relying solely on drum inspections, and even the tubes that can be seen from the drums is not by itself
sufficient to ascertain the internal boiler condition.

Visual inspections can be enhanced through the use of TV cameras and borescopes to observe the more
important heat transfer areas of the boiler. While it is rarely possible to look at the entire internal surface
area, the observation of carefully chosen areas can provide a significantly improved assessment of the
boiler’s waterside condition. Video and digital images can be recorded during these enhanced visual
inspections so that a larger audience than just those who crawled through the boiler can be part of the
decision making process. It is often necessary to look at a clean tube with these enhanced visual inspection
techniques because the devices frequently magnify what are routine tube surface effects into what appear to
be significant problems.

Boiler Contamination Events. While great efforts are made to prevent either poor quality makeup water
or steam condensate from entering the boiler, these conditions do occur in paper mills from time to time. It

Presented at TAPPI Fall Technical Conference – San Diego, CA, September 2002.
is important that the mass of potentially depositable material be considered in these cases as part of the
decision making process on the necessity to chemically the boiler following a chemistry excursion.

Another dangerous chemistry excursion is the low pH event. This can have the effect of doing an on-line
acid cleaning, which can move minor amounts of deposition from where they are tolerable in the boiler to
areas of higher heat input. Each mill should have a clearly defined procedure for dealing with boiler water
low (acidic) pH events. The response needs to be much more than a matter of dosing caustic in an attempt
to escape a potentially corrosive chemistry condition. On many occasion the feed of caustic has caused
iron that had been solubilized into the recirculating boiler water to precipitate and cause serious deposition.
Depending on the severity of the boiler chemistry excursion, and the observation of excessive particulate
iron in the boiler water, a chemical cleaning may be required on this basis alone.

New or Replacement Boiler Tubing. All new boilers require some sort of chemical cleaning before they
are put into surface. This may be something as simple as an alkaline boil-out for a low pressure unit, up to
a modified acid cleaning for higher pressure systems. Operating companies should work closely with the
boiler manufacturer to take steps to minimize the use of excessive lubricants in the tube rolling process to
avoid material accumulation on the tube and drum surfaces.

An industry rule of thumb is that replacing more than 10% of the tubes in a boiler will also require some
sort of cleaning. In several cases where large replacement panels were being installed into already clean
boilers, the panels were cleaned prior to installation, externally from the boiler. This results in a significant
time savings, reducing the outage by as much as 48 hours.

CONCLUSIONS

There are several deposit weight density test methods available to the industry that can provide important
information as to the need to chemically clean a paper mill boiler. No one procedure should be considered
as right or wrong. It is important to evaluate the various test methods in that different conclusions about
the need to chemically clean the boiler can result. Any report of a deposit weight density test result should
include an identification of the test method so that the test value can be put into the proper perspective.
Some mills are gaining additional insight by using more than one DWD test method, although a larger tube
specimen may be required.

Deposit weight density test results should be considered as part of a Body of Knowledge in making the
decision about when a paper mill boiler requires chemical cleaning. There are few absolutes in the decision
making process, but there are many more situations where a number of factors need to be considered.
Experience and sound engineering judgment must be applied to maintain safe and reliable boiler systems.
There are a number of excellent consultants available to the industry who work primarily in the field of
chemical cleaning, and their expertise can be invaluable in making sure the most effective procedure is
utilized, and that necessary precautions are taken before a cleaning contractor is put to work.

Presented at TAPPI Fall Technical Conference – San Diego, CA, September 2002.