Cement Industry Environmental Consortium

A Study of the NOx Reduction Potential of Municipal

Sewage-Derived Biosolids

Eric G. Eddings, Dana Overacker and Chris Merrill

May 5, 2003

The Cement Industry Environmental Consortium (CIEC) is proposing to use biosolids

injection as a NOx control strategy for cement kilns. The injected biosolids are presumed to

yield ammonia-like compounds that could serve as SNCR reagents when introduced at the

appropriate temperatures. The Combustion Research Group in the Department of Chemical

and Fuels Engineering at the University of Utah was contacted to perform some preliminary

testing of the efficacy of the biosolids for NOx reduction. The following provides a report

of these preliminary investigations.

Evaluation of the NOx reduction capability of the biosolid material was performed in

a bench-scale test furnace, shown in Figure 1, which operates at approximately 120,000

Btu/hr. This multifuel combustion research facility has a U-shaped configuration and will

hereafter be referred to as the U furnace. The U-furnace is a down-fired, refractory lined

reactor with an internal diameter of 6". The firing rate for all biosolids testing was 100,000

btu/hr through a premixed natural gas burner. Natural Gas was utilized as the primary fuel

in the main burner to provide suitable flue gas, NO emissions and temperatures for

subsequent downstream biosolids injection. Exhaust gases were analyzed using a

Yokogawa Zirconia Oxide in-situ oxygen analyzer, a chemiluminescent-based NO analyzer

from Thermo Electron, and an IR-based CO analyzer from California Analytical.

Materials to be evaluated were fed through a K-tron Soder dual-auger feeder. From

the feeder it was conveyed pneumatically with air to the different injection ports

investigated.

Several different materials were provided by CIEC for testing, and these included:

- dried biosolids

- wet biosolids

- rice hull ash

- biomass ash

The wet biosolids material provided had a solids content of 23% by weight. Some of

these materials were to be investigated individually, specifically the dried biosolids and the

two ash samples, to determine any NO reduction capabilities. In addition, the wet biosolids

were to be mixed with each of the two ash samples at a mixture level specified by CIEC,

and then the resulting mixture tested for NO reduction potential. The mixture level utilized

was 13% wet biosolids slurry and 87% ash material by weight. Because the wet biosolids

slurry was only 23% solids, the actual biosolids content for each of the final wet

biosolids/ash mixtures was 3%.

The furnace was operated initially with natural gas in the main burner and no

biosolids injection. Temperature measurements were then made along the axis of flow to

determine suitable injections points to facilitate reduction of NO by any ammonia evolved.

Known temperature windows for Selective Non-Catalytic Reduction (SNCR) by ammonia,

(first sampling port in Section 3) and Port 4-1 (first sampling port in Section 4) had

temperatures of 1793 deg F and 1600 deg F respectively, and were utilized for most of the

biosolids injection tests. Some preliminary tests were carried out in Port 2-2 (second

sampling port in Section 2), where the temperature was approximately 1850 F.

In addition to temperature, the other criteria used to determine the optimal injection

location during the subsequent biosolids injection tests, were the level of CO in the exhaust

and the appearance of the ash in the final filter. These additional parameters would provide

an indication of the destruction efficiency of the organic biosolids material.

The conditions in the furnace were a firing rate of 100,000 btu/hr on natural gas with

gas on a dry basis). As the test materials were fed in, the air for biosolids transport was

minimized to a level just above the point where clogging of the feed lines would occur.

In order to determine the amount of ammonia evolved from the biosolids, and thus

determine an appropriate mass flowrate for our testing, some preliminary pyrolysis studies

were performed on the dried biosolids. These tests were performed in a manner similar to

evolved ammonia being quantified by FTIR.

These tests yielded an evolution curve as shown in Figure 2, and by integrating this

curve and normalizing by the sample weight, a mass of ammonia evolved per unit mass of

dried biosolids could be obtained. The value obtained was 0.0017 grams ammonia per gram

of dried biosolids. This value could then be utilized to determine a mass flowrate of dried

biosolids required to reduce the NO in combustion products from the natural gas flame in

the U furnace.

0

50

100

150

200

250

300

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Initial tests were performed by injecting dried biosolids into the U furnace with the

intent of determining the ability of this material to reduce NOx. Prior to injecting the dried

biosolids, the baseline average NO concentration from the natural gas flame was 300 ppm

and CO was less than 10 ppm (as measured at the furnace exit).

The dried biosolids were injected at Port 2-2 and the exhaust NO dropped to an

average of 250 ppm and CO was less than 10 ppm. This was a preliminary result and the

test lasted only a few minutes as the feed was very difficult to maintain. The material was

being fed at a rate of 6-8 lbs/hr, based on the preliminary ammonia pyrolysis tests. The

small amount of dried biosolids material was exhausted after a relatively short period of

time.

The feedrate of dried biosolids utilized should have provided significantly greater NO

reduction than the 17% or so measured in this test. The baseline gas temperature at the point

of injection was approximately 1850 F, which should be a reasonable temperature for

SNCR; however, we are adding combustible fuel as well as ammonia with this injection and

thus the local temperatures may be higher than our baseline measurements due to

combustion of the biosolids. Thus, at the higher temperature we may have been oxidizing

some of the evolved ammonia to NO such that the net reduction would not be as high as

anticipated.

Subsequent testing with other materials utilized injection ports further downstream

where temperatures would be lower; however, there were no more dried biosolids so we

could not perform testing with this material at the lower temperatures

The rice hull ash was evaluated as to its NOx reduction capability prior to mixing it

with wet biosolids. The material was fed to Port 3-2 (1793 F) at an avg. feed rate of 7 lb/hr

and 7 % transport air (equivalent to 7% of the stoichiometric requirements) there was no

reduction in NO concentration and little increase in CO. The final filter was light gray,

indicating the presence of some carbon residue. Thus, when the rice hull ash was fed under

conditions appropriate for SNCR, there did not appear to be any appreciable reduction of

NO.

A mixture was made up of rice hull ash and wet biosolids that was 13% wet biosolids

slurry (itself 23% solids) by weight. It was "tumbled" in a closed container for several hours

to ensure a uniform mixture. A sample of this mixture was subjected to the ammonia

testing yielded an ammonia evolution for the mixture of 0.0013 grams ammonia per gram

of rice hull ash/wet biosolids mixture. This value was then utilized to determine a flowrate

for injecting the mixture in the U furnace tests.

The mixture was injected at Port 3-2 (1793 F) and the exhaust NO concentration

dropped from 273 ppm to 208 ppm, or an NO reduction of 24%. The CO concentration only

increased a small amount, from 4 ppm to 13 ppm. Injection further downstream at a

temperature of 1600 F yielded an NO reduction of only 20% and the CO increased from 4

ppm to 42 ppm. Both tests were run with a solids feedrate between 8.8 and 9.0 lbs/hr, and

the transport air used was 7% of the overall stoichiometric requirement.

The biomass ash was evaluated as to its NOx reduction capability prior to mixing it

with wet biosolids. The biomass ash was injected at Port 3-2 (1793 F ) at feedrates varying

from 4.8 to 8.0 lbs/hr at which point it began to clog up the feeder system. Under these

conditions, the NO and CO concentrations did not change more than a few percent, which is

within the noise of the baseline concentrations when firing natural gas alone. Thus, it does

not appear that the biomass ash provides any appreciable reduction of NO.

A mixture was made up of biomass ash and wet biosolids that was initially 13% wet

biosolids slurry by weight. It was also "tumbled" in a closed container for several hours to

ensure a uniform mixture. In subsequent U furnace testing, significant problems were

encountered in feeding this material due to its moisture content. Thus, the material was left

in the hood to dry out somewhat over night and injection was attempted again the next day.

This dryer material was successfully fed into the U furnace.

A sample of both the initial wet mixture, and the dried material were subjected to an

two. The results of these tests indicated an ammonia evolution for the wet mixture of

0.0013 grams ammonia per gram of biomass ash/wet biosolids mixture, and a value of

0.0004 for the dry mixture. Thus, it appears that approximately 70% of the ammonia in the

initial wet mixture was lost during the room-temperature drying process.

As was mentioned previously, initial attempts at feeding the wet mixture of biomass

ash and wet biosolids were unsuccessful. The material was allowed to sit out overnight in a

hood and the drier material was successfully utilized the next day.

The biomass ash/wet biosolids mixture was injected at Port 3-2 (1793 F). The mixture

was fed over a range of 6.5 to 9.7 lb/hr with no appreciable effect on the NO or CO

concentrations. The lack of NO reduction was undoubtedly due to the loss of ammonia from

the mixture during the overnight drying process.

A key question that must be addressed is whether or not the mixing of biosolids

slurries with the ash materials provides some beneficial effect with regard to ammonia

evolution. If the combination of the biosolids with the ash materials liberates more

ammonia than simple injection of the wet biosolids alone, then utilizing the mixture would

provide greater potential for NO reduction.

The ammonia pyrolysis tests that were performed facilitated an evaluation of whether

or not combining the wet biosolids with ash material yielded any enhanced ammonia

evolution. Although an ammonia pyrolysis test was not performed with the wet biosolids

slurry alone, pyrolysis tests were performed with the dried biosolids and with several of the

biosolids/ash mixtures. A comparison of the normalized ammonia evolution values (g

ammonia per g of material) indicated a similar order of magnitude of ammonia evolution

for the dried biosolids and for the biosolids/ash mixtures. This result would indicate that

there was little impact of mixing the ash materials with the biosolids slurry on ammonia

evolution. However, if the ammonia evolution is normalized to take into account the

amount of biosolids actually present in each sample (g ammonia per g of biosolids present

in the solid mixture), then the biosolids/ash mixtures appear to yield an order of magnitude

greater ammonia evolution. This effect is clear due to relatively small amounts of actual

biosolids in the biosolids/ash mixtures (on the order of 3%). Thus, it appears that mixing the

biosolids with the different ash materials results in increased ammonia release.

The best material to work with for both handling and NOx reduction seems to be the

mixture of the wet biosolid slurry (13 wt%) with the rice hull ash (87 wt%). Reasonable

NOx reductions of 20-30% were obtained, and even greater reduction may be achievable if

more material could be fed. Our feeder system limited the amount of solids that could be

injected and thus we were not able to evaluate higher feedrates. Still, with only 13%

biosolids slurry (by weight) in the solid mixture, and given that the slurry is mostly water

(23% solids in the slurry), the amount of actual ammonia-producing biosolids injected was

relatively minor at even our highest solids feedrates and yet we still saw very clear

reductions. Thus, greater solids feedrates should provide additional reduction. The trade-off

would come with the potential for higher CO emissions, carbon levels, or even ammonia

slip at high flowrates of biosolids injection. An optimal injection rate would clearly need to

be determined, and this would most suitably be done at full-scale where the appropriate

time-temperature profile and particulate loading could be utilized.

The biomass ash was also tested with the wet biosolids slurry, but the moisture

content arising from the same ratio of biomass ash (87%) and biosolids slurry (13%) proved

to be problematic. This material does not appear to absorb the water as readily as the rice

hull ash. We could not get the wet material to feed properly through our two-screw auger

feeder. The material was left out overnight to dry, and then it fed quite well the next day.

Unfortunately we did not see any NOx reduction with this somewhat dried mixture. There

was some anecdotal evidence mentioned by CIEC regarding the ease of the biomass ash

material to liberate ammonia (a reasonable possibility if the ash is more basic), and thus we

ran ammonia pyrolysis experiments to determine ammonia evolution for the wet and dry

materials. The "wet" or fresh material exhibited an ammonia evolution level very similar to

the rice hull ash/bioslurry mixture; however, the material that had dried out overnight

showed a drastic reduction (approximately 70%) in ammonia evolution. Thus, the material

that we were able to feed had undoubtedly lost most of its ammonia and thus was not

particularly effective in reducing NO.

These results highlight an interesting contrast between the two ash materials. The

ammonia evolution from the rice hull ash/bioslurry mix did not appear to decrease

appreciably as the material dried out, whereas the ammonia evolution from the biomass

ash/bioslurry mix clearly did. Thus, the latent ammonia in the ash/bioslurry mixtures

appears to be much more stable when using the rice hull ash versus the biomass ash. If the

ash/bioslurry combination would have to be mixed and then stored, the biomass ash would

likely cause the ammonia to be liberated over time; whereas, the rice hull ash does not

appear to do so in our limited testing.

Thus, if the biomass ash is the desired material for use, it is recommended that mixing

of ash with the bioslurry take place immediately prior to injection into the kiln (or preheater

tower). As mentioned previously, however, the handling of this mixture was quite difficult

in our feed system but perhaps individuals within CIEC have sufficient experience working

with these mixtures to be to provide a suitable method for feeding the material. Although

we did not successfully demonstrate NOx reduction with the biomass ash/bioslurry mix

directly, the ammonia evolution level of the "wet" mix in our pyrolysis tests was almost

identical to that of the rice hull ash/bioslurry mix, which we could feed and which clearly

demonstrated NOx reduction. Thus, it is anticipated that the biomass ash/bioslurry mixture

could provide suitable NOx reduction if injected at the appropriate conditions shortly after

mixing, and if the material handling issues were addressed.

Finally, there is evidence to suggest that mixing the wet biosolids material with the

ash materials tested (rice hull ash and biomass ash) has a beneficial effect with regard to

NO reduction, due to an increase in the amount of ammonia evolved (per mass of dry

biosolids) under pyrolysis conditions when using biosolids/ash mixtures versus dried

biosolids alone.

1- Merrill, C.J. and E.G. Eddings, "Characterization Studies of the Evolution of Hydrogen

Cyanide from Gold-Mining Leachate Waste," Final Report to the Cement Industry

Environmental Consortium, March 2002.

2- Merrill, C.J. and E.G. Eddings, "The Evolution of Hydrogen Cyanide and Ammonia

from Potliner and Carbide Brick Wastes for Reduction of NOx in Cement Kilns," Final

Report to the Cement Industry Environmental Consortium, January 2003.