Click for next page ( 16


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 15
Approaches to Obtaining Mass Balance Information INTRODUCTION Two general approaches are used to obtain mass balance information: the traditional EMB and the MA. Both approaches are described and compared in this chapter as a basis for discussing the potential of mass balance information for the applications described in SARA Section 313~1~. A quali- tative comparison is made from the perspec- tive of a reporting facility between the two approaches concerning amounts of data, technical skills of personnel to obtain and evaluate the data, and the resulting costs. (Only anecdotal quantitative cost data were available to the committee for its analysis.) The components of the SARA mass bal- ance definition fit closely with the tradi- tional EMB approach. However, the goals of a mass balance analysis also could be some- times met through information collection using MA. ENGINEERING MASS BALANCE The EMB concept is more than a century old. It has been the subject of numerous books (e.g., Felder and Rousseau, 1978) and is an established tool in engineering practice (Austin, 1984; Perry et al., 1984~. The basic 15 goal of EMB is closure when all chemical inputs to a manufacturing facility, outputs from that facility, and accumulations within have been identified and the masses have been measured. The mass of inputs should equalizer closely approximate the mass of outputs plus accumulations (i.e., mass balance requires "closure"~. Lack of closure indicates that errors have been made in quantifying one or more of the mass balance components or that additional information is needed. However, the errors inherent in any sampling and analysis procedure make attainment of complete closure an unrealistic expectation. All routes or streams for each process step (e.g., distillation) within a facil- ity must be identified and evaluated in- dependently to account for all chemicals entering or leaving a facility. This is a rig- orous evaluation procedure requiring ac- curate and precise measurements, and expert knowledge is needed to use it. Standard Theory and Practice of EMB Although EMB does not necessarily track specific chemicals, the committee evaluated chemical-specific applications of EMB, because SARA Section 313 requires report

OCR for page 15
16 ing releases of specific chemicals. These chemicals enter and exit processes (e.g., blending) within a facility through material streams conducted by pipes, air ducts, bar- rels, and so on. The chemicals within streams can become partitioned into gas, liquid, or solid phases (and phase mixtures), depending on the thermodynamic conditions within the streams. In addition, in many processes, reactions occur that convert input chemicals to entirely different output chemi- cals. In such cases, knowledge of the chemi- cal transformations is needed to relate the reacted inputs to the resulting outputs. A further discussion of the implications of such reactive systems for performing a mass balance evaluation is provided in Chapter 5. EMB uses measurement data and judg- ment (McMichael, 1988~. The measurements require knowledge of conservation of mass and of the standard units for expressing mass data as well as flow rates, determinations of chemical mass, accumulation, and the com- positions and concentrations of specific chemicals within streams flowing through a facility. Streams can be multiphase and chemically unstable, and they can contain components that interfere with analytical procedures (Rohlik, 1986~. Engineering judgment must be exer- cised in selecting the methods for sampling and analysis and in choosing the frequency and duration of data acquisitions on the mass flows within and across facility boundaries. Sampling procedures can range from in situ continuous stream sampling and analysis to instantaneous grab samples at periodic inter- vals. Judgment is needed to select the ap- propriate analytical procedure and to collect samples in a manner that adequately ac- counts for daily and longer-period variabil- ity in the concentration of a specific chemi- cal discharged from a facility. Temporal variation in waste-water concentration, for example, can be attributed to normal process variations in raw material content and reac- tion efficiency (Tischler, 1988~. An annual average EMB requires simul- taneous evaluation of the variations in chemical concentration and flow rate for the entire year. For a simple EMB on one chemical, such as purchased chlorofluoro- carbons used as blowing agents to manu- facture open-cell polymeric foams, variabil- ity of mass flow through the facility is manageable. In this case, essentially the M45S BALANCE INFORMA TION entire amount of chlorofluorocarbon is re- leased into the atmosphere within several hours after the manufacturing process. Therefore, the mass of chemical input to this facility would be equivalent to the amount released into the atmosphere (assuming in- significant amounts of the chemical remain in the foam product). EMB generally is difficult to conduct, however, because the number of necessary judgments tends to increase with the com- plexity of the facility, and because many manufacturing facilities have multiple unit- process components with multiple-process streams. EMB requires an evaluation of each stream, and the total flow rate and the con- centration of specific chemicals can vary significantly with the inevitable fluctuations in facility operation. The cost of analytical work required to obtain this information usually is very high, and the paperwork bur- den in collecting the information would be formidable. A simplified example of an EMB applied to all material entering and exiting an adipic acid manufacturing facility is shown in Fig- ure 2.1 (Nickolaus, 1988~. This EMB ex- ample is not chemical-specific and probably involves less-complex analytical methods than would be required if the analysis were performed for a specific chemical. This schematic also does not depict the composi- tions of streams flowing into and out of pro- cess components within the facility. In Figure 2.1, no data were available for the amount of reaction water generated from this process; it was therefore assumed that all of the water theoretically predicted was gen- erated as a byproduct from the reactor. Measurement Error in EMB Applications Several factors limit the certainty of data acquired through EMB, including facility size and complexity, instrument accuracy and application, system variability, errors inherent in sampling and analytical proce- cures, recorOkeeping errors, and waste ac- countability (McMichael, 1988~. The mag- nitude of errors in this information can vary significantly as a function of the types of industry, facility, and technology. For ex- ample, analysis to determine the annual quantity of TRI-listed contaminants within

OCR for page 15
- o a, a, o - ~ o a, . - 0 IL. O - ce ce ~ o o - Hi Q Cal . a, - A C ~ At, O ~ 48 Q \ ~ / o ._ - - CO _ ~ I C,' :' o ~ ~ , - :, - - . ~- 0 o ~ In o, tD ~ C ~o 0 o o ., t~1 l i. .~ WL -- 1 _ ~r ~ ~ ~ -- 0 ~ o e_ ~ 0 Ct ~ CO I- o ' Z cn ~ a a' ~ ~ _ _ 0 ~r 4~ tt, {~-- ~ 1 ~ CO 0 CO 0 ~ 0 ~ 0 {c _ - X :~ O .~ 1 _ V ~ ~ 17 - C ~ ~ ~ C~ 0 o _ CO ~ o o CO O o~ C'~ N ~ - 0~ - ev - - ~oY - ~ 0 ~ e - o "! .Q - z ~ 0 ~ - ~- ~ ~0 o o = ~ o o - ~ c) ~ o' ~ 0 4, '= ~c c~ ~ t' 0 ~ ' o 3-- a, o- o C ~ ~- ~ i_ ._ . ~ ~ ~o 0 O ~ ~ ~ <: cn ~_ ~: 3 03 C ~Ct _ - ' O C~ CO O CO ~ t_ ~ O O 0) t~ CD - 0~ - _ ~ _ _ ~Ct o C) 0 ._ _ ._ m IC ~ C~ ~C _ C ~ C ~ e CO Co O -. ~ C Z 0 to ~D - C C ', _ ~- 0 X O . _ 41, ~ - O ~ _ O O ~, ~_ :E o ~ Z ~

OCR for page 15
18 the copper ore fed to a smelter facility is more prone to error than is determining the annual quantity of chlorofluorocarbon shipped to a foam-blowing facility, because the copper ore composition is much more heterogeneous and variable than that of the chiorofluorocarbon feed. The errors in the data acquired through EMB make it virtually impossible to achieve closure of EMB for a manufacturing facility other than one composed of a single small production unit. Even if 99% closure were reached, a large amount of chemical mass would remain unaccounted for, especially at facilities where millions of pounds of chemi- cals per day are produced or used. The mass of a chemical flowing through a facility often is not measured directly but instead is calculated from independent measurements, such as total flow rate and chemical concentration within each stream. Mass data thus are obtained by multiplying the total flow rate (e.g., pounds of stream per day) by the concentration of each chem- ica1 within the stream (e.g., pounds of ben- zene per million pounds of stream). Errors can be introduced into the mass balance results from flow rate measurement error and as a result of the precision and accuracy limits of the analytical procedures employed to measure chemical concentrations (Klusek et al., 1983~. Sampling methods and analyti- cal procedures have greatly improved in the past decade for many toxic chemicals of concern and could help reduce measurement error. However, these advances can be very costly. For example, the use of gas chrom- atography for EMB measurements has been estimated to require an annual cost of $15,000 per stream at a multistream adipic acid manufacturing facility (Nickolaus, 1988). For petroleum refineries and some other industries, material flow rates are measured by volume instead of by mass. This adds uncertainty to EMB data because the mea- surement units must be converted. For ex- ample, crude oil is purchased by volume; to convert the volume to mass requires that the crude oil gravity and temperature at delivery be determined accurately. Furthermore, the amount of sediment and water in suspension must be known. The combined error in measuring volume and converting it to mass can be 0.3-0.4%. The magnitude of this error approaches that of the overall quantity MASS BALANCE INFORMATION of releases from a refinery, estimated to be 0.5 weight-percent of the crude oil input (T. Yosie, American Petroleum Institute, per- sonal communication, August 29, 1988~. The following example illustrates the significance of how an apparently slight de- gree of uncertainty affects EMB. An ethylene production facility that incorporates more than 200 state-of-the-art monitoring points in its system reports measurement accuracies that range from +0.25% to +1.0%, and closure to within 1% for large volumes of ethylene. When these overall uncertainties are based on daily balances and applied to the facility's current feed rates, they represent an imbalance of +50,000 lb of ethylene per day. This figure stands in striking contrast to the emissions from this facility, which are estimated to be 191 lb of ethylene per day. A monitoring program was used to confirm that emissions are not underestimated by the calculation procedure (Chlapek, 1988~. A high degree of closure could also be the result of two or more errors negating one another on EMB results. For example, errors in measuring the total amount of a volatile chemical in a stack emission coupled with an error in determining the total amount of that same chemical in the feedstock could result in an apparently accurate balance (Kaakinen and Jorden, 1973~. EMB is limited by analytical measure- ment error and lack of precise knowledge of chemical process science. For a facility in which a specific chemical undergoes no chemical change and is transferred to only one stream, EMB results are reasonably ac- curate. However, the processes in many facilities required to report under SARA Section 313 are not that simple. Many listed chemicals flowing through the reporting facilities occur in multiple phases and are routed through multiple process streams. MATERIALS ACCOUNTING PRACTICES OF POTENTIAL UTILITY MA is a means of obtaining mass bal- ance information that uses readily available information on material flow across facility boundaries. Unlike EMB, MA does not require closure and therefore does not re- quire detailed measurements of all process streams within facility boundaries. How

OCR for page 15
APPROACHES TO OBTAINING MASS BAL4NCE INFORMATION ever, MA can include measured data; it can include any or all of the SARA mass balance data that are obtained for EMB. Because MA and EMB can obtain similar types of information, the prime distinction between the two approaches is the method by which the information is obtained. MA draws on the basic accounting and production information of a facility. Such information may include purchase and pro- duction records; sales and shipping records; inventories of raw materials, intermediates (materials in process), and products; permit records of waste streams, and manifest and waste facility records. In comparison with EMB, MA is less ac- curate and less precise; however, this does not adversely affect MA's potential utility MA provides an approximation of the quan- tities of chemicals flowing through a facility for any or all of the same mass balance com- ponents that are required for an EMB. Although the MA approach generally uses information about the quantities crossing facility boundaries, its potential application to assess waste-reduction efficiency (Chapter 5) would require MA data collected from production units within ~ facility. A facility that produces several products might contain several production units each dedicated to a single product, or it might contain one pro- duction unit used to produce multiple prod- ucts. Therefore, MA for multiproduction unit facilities uses data on chemical quan- tities crossing the production unit boundaries within a facility. An illustration of points in a manu- facturing facility that can be represented by MA data is shown in Figure 2.2, which was adapted from a case study used in an indus- trial waste audit and reduction manual (Ontario Waste Management Corp., 1987~. This steel-pickling operation is used within a facility that manufactures nuts and bolts for the automobile industry. Before the steel can be drawn into wire for manufacturing, oxidized iron scale first must be removed from the steel with sulfuric acid. (Sulfuric acid is a TRI-listed chemical, and no quan- tities of the acid released in waste water should be reported if the acid is neutralized to pH 6-9, or if the acid is not transformed into another reportable substance.) Because the sulfuric acid is transformed into sulfate compounds as it flows through the pickling operation, MA can be applied more effec 19 tively by an approximate accounting of the sulfate flowing through the pickling opera- tion based on purchase and waste disposal and treatment records. The facility's purchase records provide the annual usage of sulfuric acid from which the sulfate input can be calculated. Waste disposal records indicate the amounts of spent pickle liquor and sludge (ferrous sulfate) found in the bottom of the pickling tanks. An analysis was performed to charac- terize the iron and sulfate concentrations in both wastes to determine waste-reduction or treatment options. The disposal records and the analytical results allow annual sulfate outputs to be estimated via spent liquor and sludge. Sulfate is also contained in the scrubber waste water and rinse water that are neutralized before discharge to the sewer. Sulfate releases to the atmosphere from the scrubber can be estimated from measurement data used to design the scrub- ber and the calculated removal efficiency of the scrubber. The amount of lime purchased and the portion judged to be used for neutralization during the year can be used to calculate the annual amount of sulfate output to the sewer using stoichiometric relation- ships of the sulfate-lime neutralization reaction. SUMMARY Table 2.1 summarizes the differences between EMB and MA approaches for several key parameters. The goal of closure for EMB requires extensive data and person- nel with the technical skills to obtain and evaluate the data; therefore, EMB is expen- sive. The great diversity in the facility operations and in the characteristics of the chemicals handled, however, makes cost generalizations extremely difficult. MA is less costly and complex because fewer data and less technical expertise are needed. MA is less accurate and precise than EMB, but material that cannot be ac- counted for does not necessarily reflect releases to the environment. Figure 2.3 presents an example of the difference in data requirements between EMB and MA for arsenic flowing through a primary copper smelting facility and fed into the smelter as a contaminant of the copper ore. EPA presents the mass balance approach

OCR for page 15
20 _ - ~ o ._ _ J of _ ~ Q A Ct C CL 03 o - Ct 0 o . _ 53 53 - CO n c I0 ~ Ct CO 3 D dS) D C LO 03 I, 0,~~ T 1 ~ ~ ~ Y ~- ._ o ~ ~ _ 1. -- ~' 3 of o of ._ .! ._ .4 - - - ~ ~ \ CO ~ ~ o I - of 0 C ._ 0 _ 'xs 0 ~ 0 / of y c - c - ._ ._ _ CO ._ _ ~ _ C - o Q. ~_ C~ . O - _ CO o~ - C ._ o CL - - 0 O <' C ~ 0 ~ _ ~ ~ oo et _ ~ Ct C~ et 3 o .~ 1 ~5 - ~' . . C, - c) - C~ ~ ._ o - ._ et ~C ,,, O O \ co cn \ I o CO \ ~ ~,. - - RO q~ '9R' ~D U) :' _ ._ ~_ c C) - C~ C~ - ._ o C~ c C, C~ - es o o ._ - C~ - C~ - - - 3 C: ~ LL

OCR for page 15
21 b j - - - - - - - - - - - - - - - - mu. .. = t ~ ~ ~ f; ----------------------i ~ j u, rev :c ~ E ~ o V ,_ C

OCR for page 15
22 TABLE 2.1 A Companson of Engineenng Characteristics of EMB and MA Approaches Approach Characteristic EMB MA M4SS BALAbJCE INFORMA TION Closure as objective Mandatory Not required Accurapy of data High Variable Level of detail of data Each process step Vanable (facility level acceptable) Data requirements Extensive new data Readily available Skill requirements Highly technical Moderately technical Additional cost High Low to moderate (including the use of MA data) as a poten- tial method for estimating environmental releases by taking the difference between input and product quantities (EPA, 1987~. However, EPA cautions that significant re- lease estimate errors can result from this procedure due to relatively small errors in the quantities used to estimate the releases. A properly designed and executed EMB or MA discourages and avoids the practice of assigning uncertain or erroneous data as be ing equivalences of environmental releases. -