FOCUS:

At the Harmac, B.C., pulp mill, Pope & Talbot has implemented several low cost projects to cut steam use by 100,000 lb/hr, allowing shutdown of two aging boilers

By JOHN COULSON

Pope & Talbot Cuts Steam Use, Reduces Energy Costs at Harmac Mill

John Coulson
is a process engineer at Pope and Talbot Ltd.'s Harmac mill in Nanaimo, B.C.

At Pope & Talbot Ltd.'s Harmac pulp mill in Nanaimo, B.C, two hog-fired power boilers were shut down in the spring of 1999 to re-duce maintenance and manning costs, simplify operation of the boiler house, and improve mill environmental performance. In an effort to decommission the boilers without increasing the fossil fuel consumption of the remaining boilers, mill steam consumption had to be reduced by approximately 60,000 lb/hr.

This article describes the approach used by the Harmac mill to achieve its steam consumption goals. Ultimately, the mill reduced steam consumption by 100,000 lb/hr. This reduction in steam use represented a natural gas cost savings of approximately C$6.0 million/yr. The project was done in-house, following initial assistance from a consulting engineering firm.

TRACKING STEAM CONSUMPTION. The first task was to improve tracking and reporting of the mill's steam and energy consumption. Remnants of an old reporting system existed, but it had fallen into disrepair and was not fully functional. In addition, the mill's steam distribution and flow diagrams were out of date. Along with upgrading the steam distribution network, existing flow meters were identified, calibrated, and repaired as necessary. Flow measurement deficiencies were also identified, and new meters were installed as required.

With an adequate flow measurement system now in place, the next step involved collecting and displaying the various information available from the mill's distributed control systems (DCS). A standard query language (SQL) database at the mill was already used to collect almost 2,000 online process data points, and it was expanded to include about 200 points of additional online or calculated steam and energy data.

TABLE 1. Several simple changes reduced brown stock washer
shower water use while holding washing losses constant.

The final stage of the tracking and reporting component involved developing a "tool set" to report and track the information now available in the SQL database. Excel spreadsheets were developed and linked via open database connectivity (ODBC) to the database, which then validated the steam flow information and generated reports. Daily, weekly, monthly, and yearly average and present value report formats were developed. A weekly meeting was also scheduled to discuss the reports and to raise energy conservation awareness.

PHASE 1 ELIMINATES TRAMP WATER SOURCES. While the tracking and reporting system was under development, the task of identifying steam reduction opportunities began. Previous mill investigations had identified the presence of large amounts of tramp water in the weak black liquor (WBL) system. A consulting engineering firm assisted in identifying and understanding tramp water sources and in the development of a dynamic WBL system model. Following the development of this model, it became apparent that detailed modeling would be required to further reduce steam consumption in a number of process areas. The process engineering group evaluated commercially available simulation packages to assist with this and CadSim Plusœ was ultimately purchased.

CadSim Plusœ dynamic models were developed for the evaporator/concentrator plant; the blow heat recovery system; the bleach plants; the recausticizing system; and the cleaners/pulp machines. In addition, an overall mill energy balance was constructed.

The models confirmed that excessive tramp water entering the WBL system was reducing the WBL solids feeding the pre-evaporator train. It was estimated that a series of simple modifications could increase kraft mill WBL solids from 17.0% to 18.5% solids.

The presence of tramp water in the WBL system also necessitated running the No. 3 evaporator train in five-effect operation instead of six-effect in an effort to increase the capacity of the No. 3 train. The models predicted that increasing WBL solids by 1.5% would allow the evaporators to run in six-effect mode.

It was estimated that increasing WBL solids and reconfiguring the No. 3 evaporators would save approximately 41,000 lb/hr of steam. In addition, the water required by the multiple effect evaporator (MEE) surface condensers would be reduced by almost 2,000 gpm.

Phase I results. The sources of tramp water entering the WBL system were identified and eliminated by piping changes. Sources included eye wash stations and pump seal water entering the spill collection system. Trials using waterless packing on several of the WBL transport pumps were also conducted. Unfortunately, the packing did not stand up to the service requirements and could not be used.

In addition to the elimination of obvious tramp water sources, the models helped to identify other water reduction opportunities. This included a project that allowed brown stock washer (BSW) losses to be held constant while reducing the dilution factor (Table 1). This project included re-piping the oscillating wire cleaning showers on the Cedar pulp line brown stock washers for more efficient water use. The quaternary screens on the cedar and Kamyr pulping lines were also removed, because they had become redundant due to earlier changes on the screening lines.

TABLE 3. Modifications separated the hot and warm water systems which saved energy and changed tank temperatures.

To further reduce the amount of wash water used on the BSWs, a new dilution factor control algorithm was also developed in-house and implemented. Once the WBL modifications were completed, the No. 3 evaporator set was reconfigured from five-effect to six-effect operation by removing blanks in the vapor system.

As a result of these modifications, WBL solids in-creased from 17.0% to 18.3%. This proved very close to the increase predicted by the model. The increased solids content, coupled with the reconfiguration of the No. 3 evaporator train, reduced steam use by approximately 40,000 lb/hr under summer conditions (Table 2). Actual steam savings were within 2% of those predicted by the CadSim Plusœ models. In addition, summer water consumption was reduced by more than 4,000 gpm. The project was estimated to have cost about C$150,000. Based on natural gas prices at the time, this reduction saved approximately C$2.6 million/yr for a project return in excess of 1,700%.

PHASE 2 FOCUSES ON BLEACH PLANT, RECAUSTICIZING. Phase I projects cut steam use by about 40,000 lb/hr. However, in order to decommission the power boilers, additional steam savings were necessary. As part of Paprican's Energy Cost Reduction Project (Project-496) the Harmac mill hosted an intermill course on the use of Pinch Analysis. The session was funded by Paprican and taught by SodeXpro.

Pinch Analysis1, 2 is a methodology that provides a systematic procedure to optimize heat exchanger or energy networks. The methodology includes:

• Energy recuperation (such as heat exchangers)

• Improvement of equipment energy efficiency

• Process modifications (such as effluent reduction, design changes, new operating conditions, and improved controls)

• Process/utility interfaces (such as with heat pumps, cogeneration, and mechanical vapor recompression)

The Harmac process engineering group applied Pinch techniques to the mill's heat recovery and hot and warm water systems. The results indicated that substantial energy savings could be realized by implementing two new heat recovery projects. To further validate the Pinch Analysis results, the process engineering group developed dynamic models of the proposed heat recovery modifications. The modeling work verified the findings, and, based on this, two projects were undertaken: bleach plant/blow slab heat recovery and foul condensate cooling.

TABLE 4. Phase II modifications to the heat recovery system and foul condensate cooler cut steam use significantly.

Bleach plant/blow slab heat recovery. This project was designed to decrease bleach plant steam consumption by recovering waste heat from the bleach plant effluent and by more efficient use of the energy already collected from the digester and evaporator heat recovery systems. Heat recovered from the bleach plant effluent was used in the mill warm water system. Improving the energy efficiency of the heat recovery systems involved de-coupling or separating the hot and warm water systems. The project would also lower the mill's summer effluent temperature.

The mill also undertook a series of blow slab and water system modifications. The hot water system was changed in order to keep hot water produced by the No. 4 evaporators and batch digester heat recovery system separate from the warm water produced in the rest of the steam plant and pulp mill. The hot water is used in the bleach plant hot water system and reduces the steam needed to make hot water for the plant. Following the change, water used to feed the flash heaters rose to 160ƒF compared to 140ƒF prior to the project.

At the mill's digester heat recovery and hot water system, simple piping changes were made as follows:

1. Heated water from the Kamyr condenser was rerouted to the Chemi-washer shower and wire cleaning pump suction (instead of the associated hot water tank). This change de-coupled a warm water source from a hot water source and helped heat the No. 2 hot water tank.

2. Fresh water is now used on the existing blow heat recovery spiral heat exchangers and the heated water is routed to an existing hot water tank. This change means a small amount of fresh water is heated up to 185ƒ F, rather than a large amount of warm water heated to 150ƒ F.

3. Warm water at 107ƒF is used on the pre-evaporator/concentrator surface condenser in place of 130ƒF warm water. The lower temperature results in lower water flow and thus lower volume for the same target temperature from the condensers.

In addition to piping modifications, a major revision of the blow slab control scheme was required to effectively balance and utilize the changes proposed.

The bleach/blow slab project also involved changes to the existing warm water system. Prior to the project, each of the mill's three bleach plants had its own warm water system. The project replaced this with a single, common warm water system.

Two reconditioned spiral heat exchangers were installed in the bleach plant alongside six existing exchangers that had been removed from service several years ago. The exchangers were used to exchange bleach plant alkaline filtrate with warm water from the MEE surface condensers. The warm water reaches 140ƒF after passing through the heat exchangers before it is collected in a new warm water tank. Variable speed warm water pumps and distribution piping were installed to service the bleach plants. The process changes resulted in notable temperature changes in the bleach/blow slab system (Table 3).

Total project cost was C$1.4 million. Steam savings were very close to that predicted by the model and amounted to approximately 43,000 lb/hr under summer conditions. Based on natural gas prices at that time, this saved approximately C$2.9 million/yr for a project ROI of 126%.

Foul condensate cooler project. An additional project to reduce mill steam demand called for decreasing consumption in the recausticizing area by recovering waste heat from the foul condensate. This would also reduce mill effluent temperature during the summer, which was an attractive bonus as fresh water was sewered during the summer to keep the effluent temperature at a reasonable level.

A new spiral heat exchanger was installed in the re-causticizing area to exchange mill foul condensate heat with fresh water makeup at the recaust warm water tank. In addition, the water side of the Kamyr black liquor coolers was converted from series to parallel op-eration. The warm water from the first heat exchanger was rerouted to the recaust warm water tank.

A major control strategy revision was also implemented to make efficient use of available water. Orig-inally, the water source for the recaust warm water tank was based on a split range level control strategy using evaporator combined condensate, Kamyr cold blow cooler water, and fresh water. The new strategy used the foul condensate cooler as the only source on level control and the evaporator combined condensate functions based on temperature control. The flow from the Kamyr cold blow cooler water is essentially fixed based on the temperature control strategy over the two coolers. There is no dynamic control on this source within the recaust warm water tank strategy.

The total cost of the project was approximately C$460,000. Steam savings were slightly less than that predicted by the model. This was due to difficulties in accurately determining and modeling foul condensate flow and temperature conditions from the blow heat accumulator tank. Steam savings amounted to approximately 6,800 lb/hr. Based on natural gas prices at the time, this resulted in savings of C$450,000 per year for a project payback of one year. The two projects in phase two realized significant utility savings (Table 4).

The next phase of this project will focus on water reduction. Efforts are currently under way at the mill to reduce water demand throughout the mill, which will ultimately further reduce steam demand.

The author would like to thank Doug McKenzie and Murray Walters for their assistance in the preparation of this paper.

REFERENCES

1. Noel, G. "La technologie du pinch au secours de la composante ènergètique des procèdès," J. des pâtes et papiers, Vol. 7, No. 5, Septembre/Octobre 91, p. 19.

2. Smith, G., Patel, A. "Step by step through the pinch," Chemical Engineering, Nov. 1987.

John Coulson is a process engineer at Pope and Talbot Ltd.'s Harmac mill in Nanaimo, B.C.