THERMAL PERFORMANCE EVALUATION OF MSW FIRED STEAM GENERATORS A NEW APPROACH

ROGER S. HECKLINGER AND LEONARD M. GRILLO Charles R Velzy Associates, Inc. Armonk, New York

ABSTRACT

TRADITIONAL PROCEDURES

Determination of thermal performance or ef ficiency of steam generators by traditional ASME procedures is a complex and difficult undertaking. When the fuel is municipal solid waste, additional variables are introduced to confound the proce dures further. A new approach is proposed as a basis for dis cussion. The intent of the new approach is to re duce the time and effort required to achieve agreement that a municipal solid waste fired steam generator has performed as specified and/or guaranteed.

INTRODUCTION

Few would question the need to confirm that a municipal solid waste (MSW) fired steam gener ating system can accomplish the thermal perform ance for which it was sold. Two papers presented at past ASME Solid Waste Conferences dealt with thermal efficiency of MSW fired steam generators. They will be used as examples. The performance test procedures followed for the steam generating incinerator owned and operated by the Nashville Thermal Transfer Corporation were described by Bozeka in 1976 [1] . Modifications to traditional ASME procedures were proposed by Stabenow in 1980 [2], which included example calculations. A review of traditional procedures follows for background.

For steam generating systems, the established procedure for determining that thermal per formance standards and guarantees for efficiency have been met is the ASME Performance Test Code 4.1, Steam Generating Units (PTC 4.1) [3]. Boiler efficiency by the heat loss method is defined as input minus losses divided by input. In put is determined by multiplying the weight of fuel fired times the heating value per unit weight of the fuel. Losses to be considered include [3] : 1. Heat loss due to unburned carbon in [resi due]. Par. 7.3.2.01. 2. Heat loss due to heat in dry flue gas. Par. 7.3.2.02. 3. Heat loss due to moisture in the "as fired" fuel. Par. 7.3.2.03. 4. Heat loss due to moisture from burning of hydrogen. Par. 7.3.2.04. 5. Heat loss due to moisture in the air. Par. 7.3.2.05. 6. Heat loss due to heat in atomizing steam. Par. 7.3.2.06. 7. Heat loss due to formation of carbon mon oxide. Par. 7.3.2.07. 8. Heat loss due to unburned hydrogen. Par. 7.3.2.08. 9. Heat loss due to unburned hydrocarbons. Par. 7.3.2.09. 10. Heat loss due to surface radiation and con vection. Par. 7.3.2.1o. 11. Heat loss due to radiation to ashpit, sensi-

65

ble heat in slag and, if applicable, latent heat of fu sion of slag. Par. 7.3.2.11. 12. Heat loss due to sensible heat in flue dust. Par. 7.3.2.12. 13. Heat loss due to heat in pulverizer rejects. Par. 7.3.2.13 14. Heat loss due to heat pickup by cooling water. . .Par. 7.3.2.14. Determination of the magnitude of these losses is a difficult and painstaking task. As it happens, the major heat losses in any solid fuel fired steam generator (including coal, wood and MSW) are: 1. Heat loss due to unburned carbon in residue. 2. Heat loss due to heat in dry flue gas. 3. Heat loss due to moisture in the "as fired" fuel. 4. Heat loss due to moisture from burning of hydrogen. All of the other losses together, in a well oper ated boiler, should total less .than five percent. (The exception could be a steam generator where high ash fuel is fired and the residue leaves the fur nace in a molten condition.) As examples, losses from the referenced papers are displayed in Table 1.
TABLE

(The data from Ref. [1] was not sufficiently com plete to be recast).
TABLE Losses Unburned Carbon Sensible heat in Flue Gas Heat of Vaporization Other Losses Total

2
Stabenow

[2]

3.54 11.47 12.61 3.01 30.63

percent

percent

SPECIAL PROCEDURES FOR MSW FIRED STEAM GENERATORS

ASME Performance Test Code 33, Large Incin erators (PTC 33) [4] , was developed to be used to measure performance efficiency of large MSW in cinerators. However, for the purposes of that test code, "The recovery of useful energy, or product from this process, although desirable, is not con sidered the measure of performance ," (Ref. [4] , Par. 0.1). Regardless, PTC 33, like PTC 4.1, en tails painstaking and detailed testing which may not be justified in light of the variable nature of MSW. Input is difficult to determine for MSW fired

1
Stabenow**

steam generators as the heating value per unit of weight used to determine total input is dependent on obtaining and analyzing a truly representative sample. MSW fired in mass-burning steam genera tors includes large individual pieces and great varia tion in moisture content and other fuel qualities from truckload to truckload. (When MSW is proc essed to produce a refuse derived fuel (RDF), the
percent'

Bozeka* Unburned Carbon Dry Flue Gas Moisture in Fuel Moisture from Hydrogen Other Losses Total

2.3 10.1

percent

t 5

14.5 1.2 28.1

percent

3.54 9.56 5.90 8.62 3.01 30.63

percent

problem of large individual pieces is reduced, but variations in moisture and ash content persist, making it difficult to obtain a representative sam ple. Therefore, the suggestions made in this paper could also have application where RDF is the sole or major fuel fired in a steam generator) . With the most elaborate of sampling techniques (and many have been suggested and tried), all one can learn is the heating value of the one gram sam ple ultimately fired in the laboratory calorimeter. Consequently, it is difficult to obtain agreement that the unit heating value obtained from the one gram sample truly represents the heating value of several hours' supply of MSW. With regard to losses, latent heat of vaporiza tion can exceed one-third of the total of all losses as may be seen in Table 2. This loss is solely a function of moisture content of MSW and mois-

Ref. [1] , p. 224. Note, Other Losses include 0.5 per cent "Unaccounted" . Ref. [2], p. 308. Note, Other Losses include 1.58 per cent "Unaccounted for Losses {Per Mutual Agreement)".

The major heat losses can be recast to combine moisture losses and to separate heat of vaporiza tion from sensible heat in the flue gas, i.e.: 1. Heat loss due to unburned carbon in residue. 2. Heat loss due to sensible heat in flue gas. 3. Heat loss due to heat of vaporization of moisture formed or released in combustion. As an example, the losses from Ref. [2] 'are re cast and displayed in Table 2. This was accom plished by splitting the two moisture related losses at 1,030 Btu/lb (2.40 MJ /kg) of moisture and ac counting for the remainder as sensible heat loss.

66

ture created or released in combustion of hydrogen bearing constituents of MSW. Once the MSW is placed in a furnace, the heat of vaporization loss is inescapable; and for a specific batch of MSW, the heat of vaporization loss is the same regardless of the furnace in which it is burned. When confronted with a fuel for which total heat input cannot be determined with exactitude, and a major combustion loss (Heat of vaporization) is uncontrollable, perhaps a new approach for evaluation of thermal performance should be explored.
A NEW APPROACH

new approach, like the procedures presently used, must be predicated on the system purchaser and the system supplier agreeing that the MSW ','to be fired is substantially as intended" (Ref. [4], Par. 3.5.2). That is, agreement must be reached that the MSW available for the test is not extraor dinarily wet, extraordinarily dry or of extraor dinary composition. There are three thermal performance factors that can be determined for MSW fired steam gen erating systems with reasonable accuracy. These include: 1. Firing rate. 2. Percent oxygen by volume (dry) in the flue gases. 3. Flue gas temperature. Firing rate can be determined by measuring the time required to fire a known weight of MSW while maintaining reasonably steady burning con ditions. Percent oxygen by volume (dry) in the flue gas is a function of percent excess combustion air. It can be measured and integrated while burning a known quantity of MSW in a known period of time. Measurement can be made by Orsat or by a continuous monitoring device. Flue gas temperature at a known oxygen level and a known firing rate is a measure of heat trans fer effectiveness. The temperature leaving the last heat transfer device in a steam generating system can be measured and integrated while burning a known quantity of MSW. (If the final heat trans fer device is an air heater, the temperature dif ference between air and gas is the measure of interest). A fourth thermal performance factor of con cern is more difficult to measure. That factor is unburned carbon loss, or more realistically, un-

burned combustibles loss. Unburned combusti bles leave a MSW fired steam generating system at e,ach location where ash or residue leaves the system. These locations could include the dis charge end of the grate, through the grate, boiler hoppers, economizer hoppers, air pollution control equipment hoppers and the stack. Fortunately, it is .normal practice to combine most, if not all, of these residue streams prior to removal from the plant. Rarely do more than two discrete residue streams leave a plant. For this dis cussion, it is assumed that the air pollution control equipment is sufficiently efficient that the com bustibles loss in particulates in the stack gas dis charged to the environment is small enough to be disregarded. For example, if particulate emissions from the stack are 0.1 grains per standard cubic foot corrected to 12 percent carbon dioxide and the particulates are 50 percent carbon, the heat loss is less than 0.15 percent. However, in most cases, residues leave the system in a wetted condi tion. Considering this, the following procedure is suggested as a means of determining or verifying unburned combustible loss : 1. Collect a large quantity, say a 55 gal drum, of mixed residue as it is discharged from the sys tem. 2. Spread the residue on a floor or a plastic sheet (do not be concerned about water runoff). 3. Pick out all easily identifiable metal and glass (do not be concerned with particles adher ing to the picked pieces). 4. Pick out large thick pieces of partially burned materials (such as logs or timbers). 5. Quarter the remainder ,down to a sample weight of 3 to 5 lb (1 to 2 kg). (Tear or cut par tially burned items such as magazines to main tain representativeness.) Dry and ash the sample following the method developed by Hecklinger and Large for determining combustibles content of RDF [5] . 6. The percent weight reduction of the dry sample as it is ashed is the percent of combustibles in the residue. 7. If fly ash leaves the system separately, sam ples of fly ash should be taken and analyzed for combustibles. 8. Repetitive samples would be desirable. This outline of a procedure is proposed for dis cussion purposes. The object is to evaluate the quantity of unburned material in the residue. It is not proposed as an absolute measure of unburn ed combustibles. One should not be too concerned

67

to find partially burned logs or timbers in the resi due; but one should be very concerned to find an undue quantity of unburned paper and rags in the residue. The purpose of the proposed procedure is to evaluate that quantity. Of course, any specification or required guaran tee for unburned combustible levels would have to be founded on specific methods of measurement such as the one described above. Note, the percent unburned combustibles in the residue will be con siderably greater than the percent heat loss due to unburned carbon in residue. Finally, one should not lose sight of the fact that the unburned combustibles loss is a relatively small loss compared to heat of vaporization loss and loss due to sensible heat in the flue gases.
SUMMARY A contractual relationship could be entered

conditions: 1. MSW consumed per hour. 2. Oxygen percentage in the flue gas. 3. Flue gas temperature. 4. Combustibles in residue.
S. Combustibles in fly ash (if separate from

other residue). Factors one through three should be monitored over a period of at least 6 hr commencing with steady state conditions. Several samples should be obtained over the same period for factors four and, if required, five. If contracted performance standards are met for each of these factors, it can be said that the MSW fired steam generation system has performed thermally as stipulated in the purchase contract.
DISCUSSION

into based on the following performance guaran tees to be demonstrated under normal operating

The method of evaluating thermal performance suggested in this paper has the virtue of simplicity

RELATIVE THERMAL PERFORMANCE

BETTER

\ \ \ \ \ \ \ \ \ \ \ PERCENT COMBUSTIBLES IN RESIDUE

o I I 10 1$ 20 I ,';1

it

WORSE

>f

25 I

30 I

PERCENT COMBUSTIBLES IN RESIDUE

o 5 10 15 20 25 o 30

/'

FLUE GAS TEMPERATURE

FIG.

1.

PROPOSED NOMOGRAPH

68

in that four or five, relatively easy to measure, factors are monitored over a several hour demon stration period. On the other hand, the PTe 4.1 test procedure develops only one number - per cent boiler efficiency - albeit by a complex, ar duous methodology based on numerous, difficult to measure factors. Thus, of the number of losses that make up efficiency, some may exceed ex pectations while others may fall short and still result in an overall efficiency within the level guaranteed. For example, a higher than expected heat loss due to unburned carbon in residues may be compensated by a lower than expected heat loss due to heat in dry flue gas. The methodology described in this paper, how ever, could be developed into a means of stipulat ing acceptable relative unit performance levels by using a nomograph to correlate the various factors ' suggested in this paper. Figure 1 is such a nomo graph developed as an example to correlate the performance factors for a MSW fired steam gen erator. The use of the nomograph is demonstrated as follows. The dotted line is drawn based on specified performance factors and ends on line AB. The point where the dotted line intersects line AB is an indication of overall specified performance. When actual performance factors are interconnect ed as shown by the dashed line, the point where the dashed line intersects line AB indicates that the net performance under actual conditions was better than the specified performance. Acceptable performance could be stipulated as that better

than a point on the line AB, or any intersection with line AB falling within an indicated range on the line. Note, the factors would have a degree of com pensating variability. For example, wetter MSW would tend to result in a lower flue gas tempera ture (higher efficiency) and in a higher oxygen level (lower efficiency). And, higher oxygen level (lower efficiency) would tend to result in lower combustibles in the residue (higher efficiency). The method of evaluation proposed in this paper is offered for discussion in an effort to simplify a perhaps unnecessarily complex process of performance evaluation.
REFERENCES
[1] Boseka, C. G., "Nashville Incinerator Performance

Tests," Proceedings of [21

1976

National Waste Processing

Conference, ASME, New York, N.Y., 1976, pp. 215-227. Stabenow, G., "Predicting and Testing Incinerator -:- Boiler Efficiency: A Proposed Short Form Method in Line With the ASME Test Code PTC-33," Proceedings of

1980 National
[3] 1965. [4] [5]

Waste Processing Conference, ASME, New

York, N.Y., 1980, pp. 301-313. Anonymous, "Steam Generating Units - Per formance Test Codes," PTC 4.1, ASME, New York, N.Y., Anonymous, "Large Incinerators - Performance Hecklinger, R. S., and Large, R. M., "Determina

Test Codes," PTC 33, ASME, New York, N.Y., 1978. tion of the Fuel Characteristics of Refuse-Derived Fuels by Macroanalysis," Proceedings of the Seventh Mineral W5Iste Utilization Symposium, IITRI, Chicago, Illinois, 1980, pp. 84-90.

Key Words Efficiencv Performance Refuse Steam Testing

69