Successful Rehabilitation Project for Jordan Phosphate Mines Company
Total Page:16
File Type:pdf, Size:1020Kb
SUCCESSFUL REHABILITATION PROJECT FOR JORDAN PHOSPHATE MINES COMPANY By: Paul A. Smith This report was not available at the time this book was prepared. ( PRAYON - RUPEL TECHNOLOGIES S.A. Ucensing Division rue Joseph Wauters 144, B-4480, ENGIS - BELGIUM Tel.: +32/41/73.93.41 Fax.: +32/41/75.09.09 SUCCESSFUL REHABIUTATION PROJECT FOR JORDAN PHOSPHATE MINES COMPANY AT AQABA,JORDAN by Mr. Paul A. SMITH of PRAYON - RUPEL TECHNOLOGIES, ENGIS, BELGIUM Presented at the AIChE Clearwater Conference - May 1994 4 1 SUCCESSFUL REHABIUTATlON PROJECT for JORDAN PHOSPHATE MINES COMPANY at AQABA, JORDAN Summary The Phosphoric Acid Plant commissioned in Aqaba, Jordan in 1982, based on a Dihydrate Process incorporating a Single Reactor and a Single Filter, was unable to reach its nominal capacity due to several process and equipment deficiencies. Due to the fact that this unit was an integrated part of the overall Fertilizer complex, consisting upstream of two Sulphuric Acid Plants of 1800 mtpd [1980 stpd) H2SO. and down stream DAP facilities of two lines of 1150 mtpd [1265 stpd) product each, the shortfall in acid production seriously affected the operation and consequently the financial results of the complex as a whole. In 1989 Jordan Phosphate Mines Company (JPMC), owner of the Fertilizer Complex, decided to rehabilitate the Phosphoric Acid plant to achieve the nominal capacity, product quality and yield. After reviewing several available options, JPMC chose to convert the plant to one based on the multl compartment PRAYON Mark 4 Process. In this paper the performance of the Reaction, Fiftration, Concentration and Cooling Water sections of the plant before and after rehabilitation are reviewed along with the major boHle-necks encountered during the former years of operation, along with a description of the process modifications made and the results of the performance test. The historical information on the original plant performance was provided by JPMC but additional information will be available when JPMC presents a paper in October at the IFA Technical Conference in Amman, Jordan. 2 SECTION 1 PROCESS DESCRIPTION OF ORIGINAL ( PLANT AND REVIEW OF PROBLEMS A flowsheet of the original plant design is attached as Figure 1 and a plan layout is shown in Figure 2. Phosphate Grinding The design specification for the phosphate feed to the reactor was 75% through 200#. Due to mill problems, some of the ball charge had to be removed in 1984 and from that time the ball mill has operated close to 65% through 200#, even though the throughput was lower due to capacity limitations caused by other operational problems. Reaction The single reactor, with a useful volume of 1250 m3 [330,200 US Gal], a specific volume of 3 1 m /mtpd P20 S [240 US Gal/stpd], is agitated by a single central agitator designed to incorporate the phosphate in the slurry and to homogenize the reactional mass. Screw type agitators act as surface coolers and some of them are also used to disperse the sulphuric acid through the slurry. As the available reactor surface was limiting for air cooling, additional cooling was provided by diluting the sulphuric acid down to 80% H2S04 and removing the heat of dilution in dilution coolers. The principal problems were the following .- Poor performance of the dilution coolers (low capacity & scaling) Insufficient cooling, requiring operation at a lower strength to prevent Hemihydrate formation (26-27% P20 5 ) Poor crystal formation Cake efficiency 93 - 94.4%, lower than design (95.5%) Average capacity of 60-73% of nameplate capacity. Filtration The filter selected was a UCEGO N°12 with three washes but the poor quality of the gypsum meant that the water soluble losses were much higher than expected. The design wash cycle was once every two days but the high reaction temperature and the lack of maturation volume caused severe filter scaling, resulting in the need for frequent mechanical cleaning of the cloth support grids and frequent cloth changes. 3 Evaporation The evaporator design basis was to concentrate 1250 mtpd [1375 stpd] P20S from 30 - 54% P20 S in three parallel lines. Each line was fitted with one block heat exchanger and one flash chamber. The units neither reached the design instantaneous capacity nor the design on-stream factor due to scaling problems in the blocks. In addition, the increased evaporation load due to the lower strength of weak acid aggravated the problem. The poor performance of the cooling water pond (see Cooling Water) meant that the design vacuum was difficult to obtain and this increased the problem of scaling even more. Cooling Water The cooling water return was fed to a cooling pond with surface sprays to increase efficiency. However the capacity was only 60% of the design and even at the reduced production rate the cooling water feed temperature was often as high as 45°C [113 oF] and not the 30°C [86 OF] as expected. c 4 SECTION 2 ( PRAYON PHILOSOPHY Flash-cooling The PRAYON process, since its origins, has been characterised by a multi-tank design with flash-cooling, which has proved to have greater flexibility than air-cooling especially when cooling capacities above the nominal capacity are desired. The low level flash-cooler, LLFC, was incorporated Into the Mark 4 design to enable the operation of the flash-cooler circuit without scaling and with a low energy consumption even at cooling rates above the design capacity. The high recirculation rate is ensured by a low head, high flowrate axial flow pump with a low power consumption and giving a aT of about 2°C [4°F] under design conditions. Dilution cooling Although the Mark 1 and 2 PRAYON designs had dilution of H2S04 to 50-60% H2S04 and removal of the heat of dilution in coolers, since the 1970's the PRAYON Mark 3 and subsequent Mark 4 designs have been characterised by the feeding of 98% H2S04• To attain this aim the Mark 3 design was specifically modified by the deletion of the dilution coolers and included the boosting of the flash-cooler capacity to balance the total cooling load, this also enabled the feeding of phosphate slurry. The H2 S04 dilution cooling system previously used had a very bad image and many operators of Mark 1 and 2 units were well aware of the maintenance headaches that these units have caused. Digestion Section The Mark 3 units also heralded the provision of a digestion section to ensure full de supersaturation or maturing of the slurry prior to the filtration stage thus minimising filter scaling and down time, due to less frequent washing and cleaning. The normal standard split for a PRAYON reactor Is 66% for Attack and 34% for Digestion this is shown in Figure 3. By changing the point of introduction of the sulphuric acid the volume of, or the residence time in, the "Low Sulphate Zone" can be adjusted to be larger or smaller. Obviously the original design at Aqaba did not incorporate this experience As such the proposal for the Rehabilitation of the Aqaba unit required a great deal of creativity to Incorporate the existing equipment into a flowsheet that resembled a PRAYON unit and more importantly was able to perform as well as a PRAYON unit. 5 SECTION 3 PRAYON FEATURES INCORPORATED INTO THE REHABILITATION DESIGN. The plan layout of the rehabilitation is shown in Figure 4 and the flexibility of the "Low Sulphate Zone" can be adjusted as in the PRAYON standard design, see Figure 5. Reaction Volume The reaction volume was increased, almost doubled, by the provision of five additional tanks, two in the AHack section and three in a newly created Digestion section. The final total volume was 3 3 2530 m [670,000 US Gal] giving a Specific Volume of 1.9 m /mtpd P20 5 [455 US Gal/stpd P20J. The final spilt of volumes was 67% for the AHack and 33% for the Digestion. Although the split follows the PRAYON standard the large size of the existing tank means that this is equivalent to a number of smaller compartments. An overall view of a 3-D computer layout is shown in Figure 6 - New Reaction Section. Phosphate feeding Two highly agitated square tanks were added each with a volume of 220 m3 [52,700 US Gal) both fitted with a three stage agitator, two PRAYON 4-PHT helicoidal sets of blades in the slurry and one PRAYON GTA at the surface, the agitator drawing 119 kW. The phosphate was diverted from the old reactor to the first of these tanks where it is was mixed with the cooled recycle flow from the flash -cooler. The return acid from the filter is fed to the first and/or second square tank pre-mixed in a Mixing Tee with part of the 98% H2S04• The slurry returns to the main reaction tank by two overflow channels. The layout of the New Attack Tanks is shown in Figure 7. Main Reador Very few changes were made to the main reactor except for the deletion of the dilution coolers, the balance of the 98% H2 S04 being added to the dispersers fitted to three of the surface agitators. The suction pipe to the flash-cooler was inserted into this compartment and the flow to the flash cooler is effected by the suction of the vacuum in the flash body. Flash Cooler The low level flash cooler, LLFC, of approximately 7 metres [23 ft) diameter was installed between the main reaction tank and the two additional tanks. The recirculation pump was fiHed on the outlet and provides a large flow of 10,600 m3/h [42,300 US GPM] with an axial flow pump having a TDH of less than 0.9 metre [3 ft), absorbed power 91 kW.