Features Research & Innovation
ICRC CONFERENCE COVERAGE: Novel technologies in the forefront at chemical recovery conference

From June 11 to the 14, 314 delegates from 12 different countries gathered in scenic Whistler, BC, for the seventh International Chemical Recovery Conference. Not surprising, considering the conference venue, that the attendance list was, in the words of Conference Chairman Mikko Hupa, Abo Akademi University, “enriched in Canadians” who accounted for about half of the attendees. The conference featured 77 papers, organized largely in two concurrent sessions. Thirty-six additional poster presentations were set up for viewing on the Monday afternoon. The conference Program Chairman, Andy Jones, International Paper, and his 18-member program committee put together an excellent program covering the latest in both research and commercial innovations in the chemical recovery area.

In the keynote address, George Weyerhaeuser Jr., Vice-President Technology, Weyerhaeuser, spoke on the conference theme of “changing recovery technology to meet the challenges of the pulp and paper industry”. Weyerhaeuser indicated that we are at the top of the S-shaped development curve for recovery technology. Very large investments are required for relatively small gains in efficiency. As the industry is very capital intensive and risk adverse, the recovery boiler is a giant barrier to innovation. While a recent article in Business Week suggested that consolidation may lead to less innovation in the pulp and paper industry, Weyerhaeuser believes that global competition may, in fact, accelerate change. He expressed the opinions that innovation cannot be left up to the industry’s suppliers and that the industry has been guilty of abandoning potential breakthrough technologies too quickly. He urged companies to join the central research institutes, particularly if they cannot support their own research facilities. Weyerhaeuser also expressed the belief that adoption of black liquor gasification technology is inevitable, despite the problems that they have seen with the gasifier at Weyerhaeuser’s New Bern, NC, mill. The higher power to steam production ratios possible with gasification and higher anticipated costs for electric power in the future make the case for continuing development and utilization of this technology compelling.

The Plenary Session also featured four overview papers that are well worth reading.

Peter Axegard, STFI, presented the main findings from a Swedish research program on the “Eco-cyclic Pulp Mill” with emphasis on chemical recovery and energy. Bjorn Warnqvist of AF-IPK AB, Stockholm, reviewed a technical and economic evaluation of black liquor gasification technologies, while Honghi Tran, University of Toronto, summarized 25 years of research on the variables influencing the sticky temperature of recovery boiler fireside deposits.

Jim Keiser, Oak Ridge National Laboratory, Oak Ridge, TN, reviewed what is now known about the cracking of composite tubes in recovery boiler floors and in primary air ports. Keiser compared the cracking in the two locations, reviewed the similarities and differences in the mechanisms, and presented a list of recommendations to avoid floor tube cracking. While research on primary air port cracking has not progressed to the point that a comparable series of recommendations can be made, he expressed the opinion that elimination of the temperature fluctuations that are observed at the bottom of some air ports might be sufficient to prevent cracking. This might be accomplished by altering the gas flow, either locally around air ports, or on a larger scale around the boiler, to prevent the delivery of fuel and/or air to the bottom of the primary air ports.

Recovery boiler safety

Glen Treger, Andritz-Ahlstrom, discussed recovery boiler smelt pumping, an evolving technology that removes smelt from the boiler floor, shortening the bed cool-down time required prior to water-washing or maintenance shutdowns. This process also recovers chemicals, reduces the time and effort required to hydroblast the remaining “frozen” smelt and minimizes composite floor tube cracking caused by water-washing when the floor tubes are still hot. Smelt pumping can reduce the time required for bed cooling and removal by up to 35 hours. It has been employed on 26 recovery boilers in North America and South America. The price of the smelt pumping service is based on a fixed mobilization fee and an additional performance fee based on the volume of smelt pumped out of the recovery boiler into the dissolving tank.

Tom Grace, University of Toronto, described the results of mill trials using sodium carbonate or liquid carbon dioxide to cool a char bed after a simulated emergency shutdown. Injection of these coolants below the surface of the bed — into hot areas where molten smelt and burning char are present — is the key to the cooling process. Grace found that either coolant could be used effectively with no significant problems. In almost all cases, the costs of using more chemicals through early application of the coolants to a hot burning bed are far overshadowed by the savings brought about by shortening the cooling time.

Recovery boiler plugging and

deposit chemistry

Tarja Tamminen, Abo Akademi University, Turku, Finland, described work done at the Metsa-Rauma mill and at Metsa-Botnia’s Kaskinen mill to determine the relative importance of fume produced from burning liquor droplets versus that volatilized from the char bed. The Recovery Boiler Dust Analyser (TAPPI Journal, 84:5, May 2001) was used to measure the amount and composition of flue gas dust just in front of the electrostatic precipitators in each boiler. Dust samples for particle size analyses were also collected using a 13-stage, low pressure, impactor. Combustion air flows were kept constant as the boiler load and number of liquor guns were varied. Both boilers were burning softwood liquor, although the solids concentrations in the fired liquors were different (71 to 74% in one and 80% in the other).

Tests showed that at full load the amount of fume was 6 to 8% of the black liquor solids (BLS) load on the boiler. In both boilers, the amount of fume tripled from 5 to 15 g/s m2 of floor area as the boiler heat load was increased from about 0.6 to 1.1 million Btu/hr/ft2. When the black liquor flow was interrupted, the amount of fume dropped dramatically and the average particle size was reduced. Tests with no liquor firing indicated that 90 to 95% of the fume originated from burning black liquor droplets and that only 5 to 10% of the fume came from the char bed in these two boilers.

The results suggest that optimizing liquor spray patterns to deliver dried liquor droplets primarily to the char bed will not only reduce carryover and plugging, but will also substantially reduce fuming, the boiler dust load and particulate emissions.

Saied Kochesfahani, University of Toronto, similarly reported that in-flight combustion of dried black liquor particles in its entrained flow reactor resulted in 5 to 15% by weight of the black liquor solids being converted to intermediate sized particles (ISP), 1 to 100 microns in diameter. Tests in three mill recovery boilers showed that ISP constitutes 20 to 40% of the material depositing on screen and superheater tube banks and 45 to 60% of the material depositing at the generating bank inlet. Increasing the size of the burning black liquor particles increased the quantity of ISP. The majority of the ISP appeared to form during char burning and not during the pyrolysis stage of combustion.

Reyhaneh Shenassa, Kvaerner Pulping, summarized work done at the University of Toronto to investigate the effect of the liquid content of carryover deposits on deposition in recovery boilers. Using the entrained flow reactor, she found that liquid content was the most significant parameter affecting particle deposition. A minimum liquid content was necessary for particles to be sticky. Larger particles required a higher minimum liquid content to become sticky than small particles; a 100-micron particle required only 12% liquid content to deposit at 800C while a 500-micron particle required 43% liquid content for
deposition. Under all conditions where the liquid content exceeded the minimum level, larger particles were deposited with higher efficiency than small particles due to their greater inertia.

In a companion poster presentation, Majid Fatemi, University of Toronto, showed that unburned char in carryover also had an important effect on the fouling tendency for carryover particles. As the unburned char content decreased from 6 to 2%, the particles got stickier at a given temperature and the deposition rate increased by a factor of 3 to 4 at 800 to 850C.

Evaporator condensate

treatment

Kent Sandquist, Excelentec, described a novel technology to treat contaminated condensate from black liquor evaporation and both high volume low concentration (HVLC) and low volume high concentration (LVHC) non-condensible gases (NCG) in a combined closed-loop system. The proposed process uses the HVLC gases and a small amount of air to strip NCGs from the contaminated condensate. The sulphur-rich gases from the stripper are burned in a regenerative thermal oxidation (RTO) unit. Part of the product sulphur dioxide is sent back to the stripping tower where it is used to lower the pH of the condensate, thereby increase the stripping efficiency. The rest of the sulphur dioxide is removed from the flue gas using white liquor in an SO2 scrubber. Demonstration plant tests at the Stora Enso Gruvon paperboard mill in Sweden showed that more than 99% of the sulphides could be stripped if the pH was kept around 4, even when the liquor/gas flow ratios were very high (14 kg/kg). High methanol removal efficiencies (~90 %) could be attained using low liquor/gas flow ratios.

Black liquor gasification

One complete morning session was devoted to the subject of black liquor gasification. It included updates on the commercial implementation of both the Chemrec and MTCI/StoneChem technologies, as well as results from fundamental laboratory studies on the evolution of gas phase sulphur and pilot plant tests on a pressurized, air blown gasifier. Craig Brown, Weyerhaeuser, gave an update on the operation of the Chemrec gasifier at its New Bern mill. It was operating from December 1996 until January 2000 when extensive cracking was discovered in the reactor shell. Investigations indicated that replacing all or most of the vessel was the only practical means to repair the unit. While both Weyerhaeuser and Chemrec are committed to repairing the gasifier, the unit is not expected to start back up until the second quarter of 2002.

Failure analyses indicated chloride-assisted stress corrosion cracking (SCC) of the 316 stainless steel reactor, probably caused by water vapour condensation on the interior of the vessel during start-ups, shutdowns and cold weather operations. The most likely source of chloride was found to be the mill water, which contained about 20 ppm. The refractory lined portion of the new vessel will be fabricated of carbon steel, with a corrosion allowance provided. This avoids the potential for SCC but opens the door to more general corrosion mechanisms.

Provisions are also being taken to ensure that interior condensation will not be experienced during normal operation. A key problem yet to be solved is the irreversible refractory expansion which created high stresses on the original reactor shell. A number of alternative refractories and designs are still being evaluated.

In a companion paper, Chemrec’s Ingvar Landalv outlined recent changes in the ownership of the Chemrec technology and the company’s plans for both the atmospheric “booster” technology and pressurized black liquor gasification combined cycle (BLGCC) technology. In spite of the problems at New Bern, Chemrec is still offering the booster technology on commercial terms and with performance guarantees.

A new, pressurized, oxygen-blown 20 tds/d development unit will be started up in Pitea, Sweden, by mid 2002. Studies in the pilot plant will focus on burner optimization, testing of construction materials, the quench system and a cooling screen reactor, improving the understanding of green liquor chemistry, and management of trace species and tars.

Incremental recovery boiler

capacity

Kent Sandquist described Tomlox, a patented process that uses oxygen enrichment of tertiary and/or quaternary combustion air in a conventional recovery boiler to substantially increase boiler throughput. Primary and secondary air flow rates are reduced, creating a highly sub-stoichiometric zone in the lower furnace, which increases reduction efficiencies. Reduced air flow below the liquor guns also decreases NOx formation.

Oxygen enrichment of the combustion air introduced above the liquor guns, using specially designed nozzles in the air ducting, increases the rate of combustion and gas mixing, while reducing the combustion gas volume and velocity and liquor carryover. The temperature profile in the boiler is monitored and used to re-adjust the combustion air distribution.

The process was installed on the recovery boiler at Stora Enso’s Norrsundet mill in Sweden in October 1998. While this boiler was already operating about 25% above its original nameplate capacity, tests showed that the Tomlox process allowed a 20% increase in throughput above the previous maximum capacity. Only 0.5 tonnes of oxygen were required to get one tonne of incremental pulp production. A comparison of the process costs with those for a number of other alternatives showed that the process had the lowest capital costs, about $0.02 million US per % increase in recovery boiler capacity.

Monitoring recovery system

efficiencies

Rolf Ryham, System Dynamics America, discussed how simple analytical tests, already performed in most kraft mills, can be used to determine the overall chemical recovery system performance and to establish simple rules and guidelines. He argued that mills, which vary chemical make-up widely to maintain sulphidity targets, would be better off maintaining a fixed chemical makeup charge and eliminating the process variability that causes sulphidity fluctuations.

He showed the impact of changing reduction and causticizing efficiencies on recovery system deadload, recovery boiler loading and evaporator loading. In his examples, decreasing the causticizing efficiency from 80 to 70% would increase the load on the evaporators by 0.65 tonnes of water per tonne of pulp and increase the solids load on the recovery boiler by 6%. By tracking sulphidity from the standard ABC test against a baseline S/Na2 ratio, Ryham believes that at least 90% of all chemical recovery system problems can be detected and identified.

While I found the program to be fairly well balanced, many of the representatives from pulp mills with whom I spoke had hoped for more papers from mills relating technical and operating experiences. Many of the operations people wanted more practical information that they could take back to the mill and use right away. Some felt that too many sessions were dominated by academic and supplier presentations. Most, however, liked the long coffee and lunch breaks, and the two conference receptions, which afforded good opportunities for networking and tracking down various experts to ask about specific mill problems. Some of the mill representatives indicated that they would like to see “operational issue” roundtables, like those that TAPPI has introduced in both its engineering and pulping conferences, added to the conference program.

The next International Chemical Recovery Conference is planned for June 2004, likely in the eastern US. The conference organizers will, no doubt, be looking to increase the number of mill papers and the opportunities for operations personnel to participate in the technical program, while keeping the quality of the technical program as high as that seen in the last two conferences. The organizing committee will also face the challenge of finding a location that can match the beauty of Canada’s west coast, the “sea to sky” highway and Whistler. Neither task will be easy. P&PC

Vic Uloth is with Paprican, Prince George, BC.