Experience on various evaporator systems illustrates the advantages of this crystallizer technology for difficult pulping liquor

Willamette Uses Enhanced Forced Circulation to Reach 80% Solids

Multiple evaporation system expansions at mills in Bennettsville, S.C., and Hawesville, Ky., have provided the opportunity for Willamette to work with different types of evaporation technology, including both falling film and forced circulation. At the Bennettsville mill, a relatively new generation of high solids black liquor crystallization technology–enhanced forced circulation–has been successfully applied.

Willamette’s evaporator train has been upgraded by addition of a high solids concentrator system.

ORIGINAL SYSTEM. The original evaporation plant at Bennettsville consisted of a five-effect falling film train with soap skimming and an integrated foul condensate stripper. A sixth effect was added in 1992. The system produced 53% solids and was followed by a dual heater, forced circulation high solids crystallizer (HSC) producing 68% solids. Product liquor flow from the falling film train was routinely switched to the second effect position to allow cleaning of the first effect by dilution. The system operated slightly above the design capacity of 2.6 MM lb/day b.l.s.

The first and second effect falling film units in the original system experienced fouling due to operation of the system above critical solids levels. Because no avenue existed to remove the scale from the heating surface faster than it formed, tube plugging eventually resulted and necessitated hydroblasting. To compensate for shortcomings of the falling film equipment, which was designed to produce 58% solids but only produced 53%, the forced circulation unit was pushed well beyond the design intentions. The excessive rate maintained in the HSC, at times approaching 200% of design evaporation, ultimately caused scaling in this unit.

INITIAL CAPACITY EXPANSION (1996). In order to de-bottleneck the mill and maximize the capacity of the (RDH) cooking system, a 30% increase in evaporator throughput was required. In addition, 80% solids were required in order to minimize changes to the recovery boiler. To meet increased demand, the original No. 1 train evaporation system was configured so that it processed 3.2 MM lb/day up to 50% solids. A new No. 2 train six-effect evaporation system was installed to process 1.4 MM lb/day to 54% from the third effect. The falling film first and second effects of the No. 2 train would concentrate the combined 4.6 MM lb/day from the No.1 train HSC and the No. 2 train third effect to a final solids of 80%.

However, following start up in 1996, initial performance of the system was poor. In addition, although it had not been required to meet design rates, severe fouling in the vessels operating near or above critical solids necessitated frequent boilouts. During the next year and a half, some difficult problems were encountered. Mill personnel were able to configure and operate the system to support mill production, although the system was never able to perform as designed and still required frequent boilouts. The trials and process modifications that led to the improved performance are summarized here and discussed in detail in the references.¹

NO.1 TRAIN HSC OPERATION. The HSC design includes a large retention chamber to provide a crystal inventory adequate for release of supersaturation developed as liquor re-circulates through the heat exchanger. Originally intended to operated at 70% solids, well above the critical solids point and in a region where a good crystal fraction would exist, the HSC was operating near critical solids at 50%. In this mode, the HSC experienced slow fouling that eventually required hydroblasting. The ineffectiveness of boilouts indicated the scale was at least partially non-water soluble, and analysis confirmed the majority of scale to be calcium carbonate and sodium carbonate.

FIGURE 1: Spiral Rib Turbulence enhancer increases heat transfer, thus cutting power use.

Operation of the HSC was greatly improved by bypassing liquor around the unit and raising the product solids to 70%. To accommodate the higher liquor viscosity, a vapor valve was installed downstream of the vessel to maintain higher liquor temperatures. Minimizing fouling in the first and second effect of the No.1 set alleviated some of the duty on the HSC. When operating as a crystallizer, and not exceeding design evaporation rates, the HSC no longer required boilouts. The HSC has operated for two years in this configuration without requiring a boilout, and as such, the mill does not even open the unit for inspection during the semi-annual outages.

NO. 1 TRAIN FALLING FILM OPERATION. The first and second effect falling film units of the No. 1 train also scaled and typically received one to five boils per month. The units were taken out of service ahead of schedule for hydroblasting. The mill concluded that the RDH pulping process produced liquor that exhibited calcium scaling at lower temperatures than typical kraft liquors. The first and second effects were operating at 110°C to 120°C and 93°C to 100°C, respectively.

Interestingly, the forced circulation HSC did not exhibit calcium fouling when operating at 70% solids and temperatures as high as 135°C. By removing an orifice in the vapor line from the third effect, the average operating temperatures in the front end of the train were dropped by 11°C. In addition, liquor routing was modified to provide product from the second effect position, such that the more concentrated liquor was at a lower temperature. With these modifications, the first effect required no more boils, and second effect boils were performed only twice per month. No calcium fouling was observed at the following scheduled outage.

NO. 2 TRAIN FIRST AND SECOND EFFECTS. The first and second effects of No. 2 train are falling film concentrators that were designed to operate at 80% and 65% solids, respectively. Although the units were designed as "crystallizers" with very large volumes (crystal inventory) in addition to large area / low heat fluxes and high recirculation rates (low developed supersaturation), bad fouling occurred.

Several equipment modifications were made to the first and second effects falling film concentrators in an attempt to lengthen the run time. The liquor distribution systems were modified in an attempt to avoid plugging and improper distribution. The recirculation rate was increased from 6 gpm per tube to 8 gpm per tube in the second effect. Neither of these modifications seemed to help.

Liquor routing and ash/saltcake addition was modified in an attempt to positively affect chemistry and/or physical properties of the liquor. To reduce plugging of the second effect distribution plate, it was reasoned that ash and salt cake, originally sluiced with 50% liquor, should be slurried or dissolved in the weak liquor feed. However, the scaling rate of the first and second effects actually increased with ash dissolved in the weak liquor, and the third effect also scaled. It is likely that dissolving the ash in the weak liquor resulted in a lower critical solids point, causing crystallization and scaling in the third effect. Salt cake and ash addition were returned to the 50% tank immediately upstream of the falling film concentrators, with further provisions made for better blending. These provisions included better agitation of the slurry and the ability to dissolve the ash with weak liquor.

Dissolving the ash in weak liquor prior to feeding the second effect resulted in an improved run time of greater than 9 days for this unit, with boilouts typically initiated by the need to wash the first effect. A trial was also performed routing boiler ash slurry directly to the first effect in an attempt to further improve the run time of this unit, but resulted in an increase in the fouling rate.

There appears to have been two scaling mechanisms occurring. First, distribution system plugging was causing improper distribution that led to tube starving/fouling. Better dissolving of the ash, which apparently reduced and/or destroyed large chunks of agglomerated ash, alleviated this. Unfortunately, dissolving the ash increased the mass of salt that had to be crystallized from solution. This led to an increase in the required levels of supersaturation and the chances for scaling due to the second mechanism, which involves improper release of supersaturation on the tube wall instead of on existing crystal surface area.

FIGURE 2: Two EHSC units in the first effect at Bennettsville on No.2 set now produce 80% solids.

FIGURE 3: Configuration at Hawesville mill is similar to Bennettsville system.

1998 OPERATIONS AND THE NEXT STEP. After two years of diligent investigation, trials, and equipment/operational changes, the evaporation system at Bennettsville was able to support the mill. The system was still quite difficult to operate, however, requiring a boil of the falling film concentrators approximately every 10 days. While the system was able to operate with design liquor solids throughput, it was seldom operated at the design product solids of 80%.

In 1996, Willamette also installed a falling film crystallization system at the Hawesville, Ky., mill and experienced similar fouling difficulties. By early 1998, both mills needed a final solution for their evaporation problems. The solution would have to allow operation at design solids and throughput, without the need for constant boilouts.

Key conclusions based on two years of operation included:

• An FCC (or any crystallizer) will typically operate better away from the critical solids point.
• FCC’s exhibit resistance to calcium scale, in part due to an inventory of crystal surface area for salt deposition.
• RDH pulping liquor has a lower temperature threshold for calcium scale. Operating at lower temperatures and routing liquor so that the highest concentration is not at the highest temperature can minimize calcium related fouling.
• It is not always possible to rely on adjustment of black liquor chemistry to prevent fouling. At Bennettsville, there were several competing fouling mechanisms, including mechanical fouling (distribution plate plugging), insoluble chemical/temperature related calcium scaling, and soluble scale chemical (burkeite) scaling.
• Maintaining a minimum carbonate-to-sulfate ratio will provide a positive crystallization chemistry and will minimize scale. This is because a higher percentage of the precipitation will occur as burkeite rather than sodium carbonate. However, any salt cake or ash dissolved to affect crystallization will have to be re-crystallized at some point. The more crystallization to be performed, the more likely it is to occur on heat transfer surfaces. There needs to be a balance between positive solution chemistry, adequate crystal seed availability, and crystallization duty–all of which affect the supersaturation levels within a crystallizer.
• Operation at Bennettsville illustrates the relative performance of forced circulation technology vs. falling film technology. The forced circulation unit was operating at the same solids and with the same feed as the falling film concentrator second effect. Because boiling and the resulting development and release of supersaturation (i.e., crystallization) occurs away from the heat transfer surface in a forced circulation unit, the susceptibility to fouling and chemistry issues is greatly reduced. This superior scaling resistance of the forced circulation design over falling film was exhibited at both Bennettsville and Hawesville.

BLACK LIQUOR CRYSTALLIZERS. In the spring of 1998, Willamette purchased enhanced forced circulation crystallization systems for both the Bennettsville and Hawesville mills. The enhanced forced circulation design is an evolution of conventional forced circulation technology that has been used for crystallization of liquor above critical solids for decades. Its design is very robust, typically with no need for washing or dilution. To avoid scaling difficulties at these solids, evaporation equipment must be designed as crystallizers to allow these salts to form in the bulk liquor and not on heat transfer surfaces.

An alternate design strategy allows fouling to occur, but provides a means to remove scale faster than it forms and before it negatively impacts capacity or leads to tube plugging. Quick "switching" designs employ this strategy by continuously moving multiple crystallizer bodies or chambers between product liquor and washing positions. One drawback of constantly switching arises from the fluctuations in product solids occurring every time a switch occurs. In addition, the multiple bodies or chambers are constantly cycling between hotter and colder positions, causing mechanical stresses that can lead to early equipment failure. Finally, complex piping and valves is required to facilitate the multiple liquor flow configurations.

The forced circulation design is the technology of choice when boiling at the heat transfer surface (as occurs with film type designs) is undesirable. This can be the case when the falling film unit requires very high recirculation rates and/or prohibitively large heating surface to deal with high developed and residual supersaturation. Also, while a falling film crystallizer can produce very high heat transfer rates at the lower viscosity associated with weaker liquor, film development and resulting heat transfer performance are adversely affected by high viscosity of heavy black liquor.

As with any crystallizer, the forced circulation unit is designed with sufficient retention volume for crystal growth. However, the developed supersaturation in the tube of a forced circulation unit is lower than in a falling film unit. This is because boiling within the forced circulation tube is suppressed (evaporation in a falling film tube increases developed supersaturation). In addition, temperature rise through the heater is selected to minimize any adverse temperature effects (sodium carbonate and burkeite salts exhibit an inverse solubility). The high tube velocity typically required for the forced circulation design also counters the effects of high viscosity via shear thinning and increased turbulence.

FCC’s can be supplied with a vertical or horizontal heater. The vertical heat exchanger design offers a compact footprint, an important consideration when space is limited. The horizontal design provides additional horsepower savings over the vertical design. An orifice plate is required at the entrance nozzle to the vapor body in order to suppress boiling in the vertical heater. The pressure drop created means additional TDH on the recirculation pump and translates directly to increased power requirements. With a horizontal exchanger, no orifice plate is required because the necessary back pressure to prevent boiling in the tubes is provided by the elevation of liquor above the heat exchanger.

ENHANCED BLACK LIQUOR CRYSTALLIZERS. While the forced circulation design can be the technology of choice for most situations, it historically has higher power requirements. However, an enhancement to lower power demand was developed. This technology is known as enhanced heat transfer, and employs a proprietary spiral rib type insert inside the heat exchanger tubes (Figure 1). The insert allows for a high degree of turbulence (at high viscosity) at low recirculation rates.

The results of enhanced heat transfer are smaller required heat transfer area and/or a reduction in required pumping power. A proprietary spiral rib type design has been developed and tested. The design has allowed for a 50% to 75% increase in heat transfer coefficient and a 50% reduction in power when compared to the conventional forced circulation design.

Initial work using black liquor was done with the proprietary turbulence-enhancing insert in 1993 as part of a study for the Weyerhaeuser’s Longview, Wash., mill, The inserts provided a 65% increase in heat transfer efficiency over open tubes. In short, the pilot work demonstrated high heat transfer performance and low horsepower requirements without fouling. The first commercial system was started up at Longview in 1995 and has operated beyond expectations with no fouling.²

Willamette elected to install enhanced heat transfer technology at both Bennettsville and Hawesville with the following goals:

• Maximum resistance to scale formation.
• Minimal piping and control requirements.
• Optimal steam economy.
• Lower required steam pressure.
• Low horsepower requirements.
• Minimal capital investment.

It required nine months to engineer, construct, install, and successfully start up the enhanced system. Two enhanced forced circulation crystallizers were installed in the first effect position of the No. 2 set (Figure 2). The existing falling film concentrator first and second effects were placed in parallel in the 2nd effect position. With two bodies in the second effect position, enough area is available to take one body completely off-line for washing without significantly reducing the capacity of the train.

The system started up in November of 1998. All performance warrantees were met and the operability of the evaporation plant has greatly improved. Currently, the HSC on the No. 1 train produces 74% solids, does not require boilouts, and has not been hydroblasted in two years. The new enhanced high solids crystallizers (EHSC’s) produce 80% solids, are not boiled out, and have exhibited no signs of fouling. In April of 1999, the EHSC heater units were opened during the semi-annual outage and no plugged tubes were observed.

The falling film units operating near or above critical solids are still susceptible to fouling and require weekly flushes. The falling film bodies of the No. 1 train are still operated with a 3-1-2 liquor flow configuration, with a switch to the 2-3-1 configuration eight hours per day for dilution and cleaning of the second effect on the go. During the April 1999 outage, the first and second effect had 50 to 100 tubes (< 5%) plugged apiece.

The second effect falling film concentrators of the No. 2 train produce 58% to 62% solids, and are flushed once per week with weak liquor. The boilout frequency or effectiveness is less than optimal, as the units were approximately 60% plugged at the April 1999 outage. The 3rd effect of the No.2 Train was about 25% plugged, but had not been hydroblasted for two years.

CURRENT OPERATION AT HAWESVILLE. The enhanced crystallizer equipment installed at Hawesville was started up just prior to the system at Bennettsville and performed similarly. The Hawesville equipment was modified in a manner similar to Bennettsville (Figure 3). Conventional forced circulation is at the front of one train, and switching falling film concentrators at the front of the other. The switching concentrators are switched and washed weekly. Both trains finish in enhanced forced circulation bodies. Some key observations to be noted regarding operation of the system at Hawesville are as follows:

• The switching falling film units must be washed for longer periods than originally expected in order to return to clean operation.
• An NSSC liquor stream is blended with ash and intermediate liquor prior to introduction in the HSC and falling film concentrators. If the solids content of this tank drops to the point that the ash is dissolved, rapid scaling (particularly in the falling film units) is observed.

EVAPORATION TECHNOLOGY COMPARISON. A rough economic comparison between the falling film, conventional forced circulation, and enhanced forced circulation technologies can be made based on performance of the various types of concentration equipment at Bennettsville. Table 1 illustrates the relative efficiency of the equipment. Note that this efficiency will be affected by the solids being produced, with higher solids production a being a more difficult service.

The performance of the enhanced forced circulation systems installed at Bennettsville and Hawesville illustrate the sharp contrast in operability between forced circulation and falling film crystallizer technology. In March of 1999, Willamette purchased another system for its Red River Mill in Campti, La. Forced circulation technology has historically proven to be very effective for black liquor crystallization at high solids. With the further evolution of this technology provided by turbulence enhancement, and the associated reduction in capital and operating costs, forced circulation technology becomes the clear choice for crystallization of high solids liquor. More than 20 enhanced forced circulation units have been purchased since the original system was started up in 1995.

Jim Rieke and Kent Drone are with USFilter HPD Products in Plainfield, Ill. Don Cox and Carl Clark are recovery managers with Willamette Industries in Bennettsville, S.C., and Hawesville, Ky., respectively.

REFERENCES

1. Gore, Christopher, TAPPI 1998 International Chemical Recovery Conference Proceedings, TAPPI Press, Atlanta, pp.33-39.

    

2. Rieke, J., J. Brinker and S. Bogart, Weyerhaeuser Tries New Crystallizer Design to Reduce Fouling, Cut Power Consumption, Pulp & Paper Magazine, July, 1997.

Based on a paper at the 1999 TAPPI Engineering Conference