Pilot plant tests comparing various alternative coal pipelining options

Page 1

bulk solids Volume 8, Number 6, December 1988 handling Papen, Cane Coal, Righy Thomas.

GRANULAR SLURRIES BULK SOrDs HANSUNG, Vol.8, Nob, Der. 1988

55

Pilot Plant Tests Comparing Various

Alternative Coal Pipelining Options

G.R. Rigby and A.D. Thomas, Australia

Summary undertaken in recent years. For fine coal systems improve- For ifferent options for pipelining one particular coking onstrated the potential for improvements over the Black ments in dewatering equipment and processes have demcoarare examined. These are: - 1.5 mm fine coal; - 15 mm Mesa operation [6-14]. coarse coal; - 15 mm stabilized coarse coal; and fine coal in the form of 1 to 2mm agglomerates. Test results obtained For coarse particle slurries, alternative technologies involvin a 1,600 m and 400 m x 100mm pilot facility are preing high concentration mixtures with maximum packing sented for all four coals. The effect of addition of electrolyte densities [15] and stabilized slurries incorporating additionon rheology and the resulting changes in pipeline behaviour al fines to act as a carrier for the larger particles [5, 16-20] are also reported. have been investigated. The results are analysed especially with regard to scale-up This paper reports the result of an investigation undertaken to larger diameter pipes. Finally the economic advantages to compare the slurry and pipelining characteristics for a and disadvantages of each option are briefly discussed. range of options for a particular run-of-mine coking coal.

The following options were tested: - fine coal slurry with a top size (99% passing) of 1.5 mm

1. Introduction (similar to the Black Mesa slurry) The Black Mesa slurry pipeline, which has conveyed some - coarse heterogeneous slurry with a top size of 15 mm 85 million tonnes of coal since commencing operations in - a stabilized fine/coarse coal mixture with a top size of 1970, has clearly demonstrated the technical and economic 15 mm viability of transporting large tonnages of coal over long — slurries containing 1-2 mm diameter spherical agglomdistances. This technology which utilises a fine coal slurry erates prepared from a fine coal slurry using oil agglomis basically limited to energy coal being supplied to captive eration. • generation stations or other pulverised fuel users. export market or for coking coal applications, primarily due Tha product is not generally suitable for the large scale 2. Coal Properties and Equipment to the size of the coal (1]. A 100 + run-of-mine raw coal sample from the Hunter Valley, Optimum pumping properties for slurries are achieved when NSW, Australia was used in the investigation. Table 1 the particle size is small. For these slurries, the pressure shows the chemical analysis for the sample and Table 2 gradient is low and pipe wear is almost negligible. However, gives a typical X-ray diffraction mineralogical analysis for there are numerous cases where it is either not economic the coal. Solids specific gravity was 1.75. to reduce the particle size or when the product is required Laboratory scale test work was undertaken at the BHP to be larger. Central Research Laboratories. Slurry rheological measure- Transportation of coarse particles in heterogeneous slurries ments were made using a Contraves Rheomat 30 viscomehas essentially been limited to relatively short distances [2ter. 5]. The two main problems caused by the coarser particle Pipelining trials were undertaken at the Stockton Borehole size are high pressure gradients and high pipe wear rate. Pilot Plant near Newcastle [10]. Some preliminary explora- Extensive research and development programs aimed at tory work was also done in the 50 mm diameter, 100 m long improving existing fine coal systems and expanding the pipeline loop at the Central Research Laboratories. range of slurry options for coarse coal systems have been The facilities at the Stockton Borehole Pilot Plant include a

20 t/h hammer mill, a 3.3 m long × 2.1 m diameter ball mill, a 300 m? agitated slurry holding tank and 102 mm internal

Dr. G.R. Rigby, Senior Principal Research Officer, BHP Central Research diameter pipe test loops of total length either 400m or Laboratories, P.O. Box 188, Wallsend, 2287 NSW, Australia, and Dr. A.D. 1,600 m. In addition a 50 m length of 200 mm diameter pipe Manuscript received: August 4, 1988. Thomas. Consultant, Cantwell Road, Lochinvar, 2321 NSW, Australia. was available. Pumping velocities were measured using a

663

Page 2

Slurry & freight pipeline technology Volume 8, Number 6, December 1988 solids bulk

ing differential pressure transducers. magnetic flow meter and pressure gradients measured us- Slurry density was 3. Pipeloop Test Results monitored using nuclear gamma ray density gauges with 3.1 - 1.5 mm Coal absolute values obtained from frequent samples. Slurry flow behaviour was observed through a transparent section of The particle size distribution of this slurry is shown in Fig. 1. pipe. Particle size analyses were obtained by sampling the The median size is 45 um. This makes it finer than the coal slurry at the highest velocity. The distributions reported slurry pumped in the Black Mesa pipeline which has a represent the average of two samples. Because of the large median size of around 150 um. The top size is similar in both volume of the surge tank compared with the pipe loop cases. The finer size consist in the present case is due to volume there was negligible attrition during each test. the high proportion of clays present. The fine coal slurry was prepared using the hammer mill and To prepare this slurry the run-of-mine coal was wet ground closed ball mill circuit to generate the desired particle size. in the ball mill with fresh water and a 39% concentration by For the heterogeneous slurry, the ball mill was bypassed. weight slurry produced. This was then tested in the 100 mm The stabilised slurry was prepared by firstly generating a diameter pipeloop. The test loop results are shown in Fig. 2. sufficient quantity of fine coal slurry which was stored in the A stationary bed was observed at velocities below 1.05 m/s: surge tank. Coarse coal was then added to this mixture in However this was mostly composed of granular mineral the desired proportion. particles (sand). At 0.95 m/s this bed was 5 mm deep. Even The coal-oil agglomerates used in the study were prepared at velocities as low as 0.40 m/s the coarser coal particles by firstly crushing the run-of-mine coal to a top size of were still slowly saltating along the top of a 20mm high bed 0.5 mm at a concentration of 67% by weight. Light gas oil which consisted of sand and finer coal particles. (17% by weight of dry coal) was subsequently added to the This was a very "thin", un-coagulated slurry. Viscometer slurry and the mixture circulated through the 400 m long tests revealed it to be Newtonian in nature with a viscosity 100 mm diameter pipeline loop until 1-2 mm diameter agof 2 cP. When left to stand it settles rapidly to form a glomerates had been formed. granular bed. During the formation of the agglomerates, liberation and Bench scale tests using a viscometer showed that addition separation of mineral matter contained in the run-of-mine of cations increased the slurry viscosity. This was likely to coal occurs and hence the ash content of the agglomerated improve the hydraulic behaviour. For this reason the same particles is significantly reduced. Negligible coal matter slurry was next tested after sea water had been added in remains in the aqueous phase. The ash content of the the ratio of 6 parts sea water to every 100 parts of slurry. agglomerates generated in this manner was 9.2% (dry, oil The effect of this addition of cations was dramatic. The free). Further details of the agglomeration process are given slurry became visibly more viscous and was now statically elsewhere (8, 10]. stable in the sense that when allowed to stand it settled only

very slowly and with no segregation of the coarser particles. Viscometer tests revealed the slurry now exhibited Bingham

Table 1: Chemical analysis for run-of-mine coal sample plastic-like behaviour with a yield stress of 0.52 Pa and a plastic viscosity of 5.9 cP at a slurry concentration of 39%. Moisture % a.d.b. 2.43 Fig. 2 also shows the test loop result for this slurry. The Ash % d.b. i 48.25 critical deposit velocity was lower than before, at 0.78 m/s. Mineral Matter % d.b. 51.22 At 1.5 m/s the pressure gradient was slightly less (320 Pa/m Volatile Matter % d.b. 20.72 c.f. 340). The higher viscosity of the coagulated clays results Carbon Dioxide % d.b. 1.01 in less heterogeneous behaviour giving a lower deposit Total Sulphur Pyritic Sulphur % d.b. 0.36 homogeneous fluid. velocity and pressure gradient/velocity plot more like a - Sulphate Sulphur % d.b. % d.b. 0.02 0.12 / Chlorine % d.b. 0.02 3.2 - 15mm Coal Hydrogen Carbon % d.b. 41.39 Another test involved the same coal crushed to 15 mm top Nitrogen % d.b. % d.b. 1.11 3.00 size in the hammer mill giving the size distribution as shown Oxygen Specific Energy [MJ/kg] % d.b. 3.31 tration slurry in the same pipeloop. The measured pressure in Fig. 1. This coal was then pumped as a 32% concen-

% d.b. 17.04 gradients are shown in Fig. 3. At 1.15 m/s a 25 mm bed was

observed slowly sliding along the pipe bottom in a stop/start fashion. Bed motion stopped at velocities fractionally below this. At all but the highest velocity tested

Table 2: Typical mineralogical analysis the flow was erratic with violent fluctuations in the pressure

gradient.

Siderite Quartz % d.b. 24 The pressure gradient at 1.5 m/s was 960 Pa/m. When flow Pyrite % d.b. % d.b. 1 3 was stopped the solid particles settled immediately. Although there was some free clay with the water this was Feldspars Kaolinite + Chlorite % d.b. 4 uncoagulated, there being no salt water present.

% d.b. 30

*Expandable clays Illite % d.b. % d.b. 14 24 3.3 - 15 mm Stabilized Coal Slurry * Montmorillonite group and mixed layer clays. This slurry was prepared by adding - 15mm hammer mill In this test - 15 mm coal was pumped as a stabilized slurry. 664

Page 3

solids bulk Volume 8, Number 6, December 1988 Slurry & freight pipeline technology handling

ROSIN RAMMLER GRAPH

99- 95 90 Minus 20

70

PERCENTAGE UNDERSIZE 40 50 Minus 15mm Stabilized 50 90 85 60

80

93

115m

] 99

0.063 0.125 0.250 0.500 1.00 2.00 4.00 8.00 16.0 31.5 63.0 125

SIZE mm

Fig. 1: Particle size distributions

product to an equal quantity of - 1.5 mm ball mill product Test results for concentrations of 48% and 58% are shown on a dry weight basis. The resulting size distribution is in Fig. 4. At the lower concentration a stationary bed of shown in Fig. 1. Aluminium sulphate (Alum) was also added particles appeared at velocities below 0.8 m/s. At the higher at the rate 0.43% on a dry basis. This served the same concentration deposition occurred at velocities below purpose as the sea water added to the fine coal slurries in 0.5 m/s. In both cases laminar flow is evident at all velocities that it coagulated the slurry, turning a rapidly settling slurry tested. into a stabilized, non-settling slurry. At all velocities above the deposition velocity the flow was

very smooth and steady. The flow could be readily stopped and restarted without difficulty.

2000- Legend • Un-coagulated

A Co-agulated 2000-

1000- •

SURE GRADIEN - Un-coagulated 800-

400- - Co-agulated

600-

PRESSURI (Pa. 200- 600- Stationary bed observed 25l A • PRESSURE GRADIENT (Pa/m) 1000- 200- 400- Stationary bed observed

100- 30 80- 20 4

60% 0.4 0.6 0.8 1 3 100L 0.6 0.81 2 3 4

VELOCITY (m/s) VELOCITY (m/s)

Fig. 2: Fine coal slurry results; numbers above data points indicate station- Fig. 3: Coarse coal heterogeneous slurry results ary bed height in mm

665

Page 4

Slurry & freight pipeline technology Volume 8, Number 6, December 1988 solids bulk

nonaline

2000- 4. Discussion of Pilot-Plant Results Stationary bed observed Figs. 2 to 5 exhibit the typical behaviour of the four types

of slurries. For example at 1.8 m/s the pressure gradients were:

1000 400 Pa/m for the agglomerates slurry

900 to 1,500 Pa/m for the stabilized coarse coal

600- 1,000 Pa/m for the coarse coal heterogeneous slurry.

Agglomerated slurries typically show somewhat lower pres-

PRESSURE GRADIENT (Pa/m) 800- 100 400- 200- 0.4 • 48% Legend A 58% 0.6 0.8 1 Water 3 4 pronounced in larger pipe sizes. is desirable to reduce pipe wear. at much lower velocities is possible. Low velocity operation slurries. 440 Pa/m for the fine coal slurry that the solids concentration is higher and stable operation size pipe. The stabilized slurry has advantages however in significantly lower pressure gradients than the coarse coal The stabilized coarse coal slurry requires similar pressure gradients to the heterogeneous coarse coal slurry in this The differences between the slurry types becomes more sure gradients than fine coal slurries. Both of them require

VELOCITY (m/s) For the fine coal and the agglomerated coal slurries the

Fig. 4: Coarse coal stabilized slurry results pressure gradient varies approximately inversely with pipe diameter. Furthermore, the deposit velocity increases only

slightly with pipe diameter. These two factors mean that in large pipe sizes the pressure gradient is low and the slurry

3.4 Agglomerated Coal Slurry can be pumped at the most economic velocity, typically 1.5 Fig. 5 shows results for an agglomerated slurry consisting to 2 m/s [21, 22]. of agglomerates of average size 1.5 mm in a clay/water In contrast, the heterogeneous coarse coal slurry requires vehicle. A typical size distribution is shown in Fig. 1. Tranhigh velocities in large pipe sizes. This is because the sition between laminar and turbulent flow occurred at deposit velocity increases approximately with the square 1.8 m/s. Total solids concentration was 57%. The agglomroot of pipe diameter. Pressure gradients do not decrease eration process causes all of the coal matter to report to the significantly with an increase in pipe size. agglomerates with the majority of the clays remaining with Scale-up of stabilized coarse coal slurry data to large pipe the water in the vehicle portion. The rheology of this vehicle sizes is complicated. If the slurry is viscous enough scaleportion was measured and it was found to have Bingham up can be performed as if the slurry were a homogeneous plastic properties with a yield stress of 1.75 Pa and a plastic non-Newtonian fluid. In this case the pressure gradient will viscosity of 4 cP. be approximately inversely proportional to pipe size, and

pumping pressures will be low. However, for coarser par-

2000eventually reached when this scale-up approach no longer. ticles and less viscous slurry combinations a stage is

applies (17]

(Pa/M) 1000- 800- 5. Scale-Up to 3 Million t/a 5.1 General

GRADIENT SURE 400- 600cular case of transporting 3 million t/a was chosen as being a better comparison of the four different options. The partipipe, can be scaled up to larger pipe sizes. This will allow The data of Figs. 2 to 5, obtained in the 100mm diameter of interest for this mine at the time. It is also a typical

PRES quantity for the Australian export coal industry.

200- Water 5.2 - 1.5 mm Coal

The test data are for a relatively low concentration of 39%. Consideration of the rheology of this slurry indicates that concentrations up to 50% could be handled. At this concen-

100 0.4 0.6 0.8 1 3 4 5 tration the pressure gradient will be approximately 35% higher than at the 39% concentration tested. VELOCITY (m/s) Using a 350mm pipe the flow velocity will be around

1.55 m/s which is a suitable pumping velocity. The data of

Fig. 5: Agglomerated coal slurry results Fig. 2 show this slurry behaving as a pseudo-homogeneous 666

Page 5

bulk andling Slurry & freight pipeline technology solids Volume 8, Number 6, December 1988

fluid at 1.55 m/s. The pressure gradient will therefore vary Table 3: Transport of 3 million t/a pressure gradient at 1.55 m/s in a 350 mm pipe is therefore Concen- Pressure roughly as pipe diameter to the power -1.15 [22). The 95 Pa/m at 39% concentration and 115 Pa/m at 50% con- Option tration Diameter Velocity Gradient centration. [%] (mm] [m/s] [kPa/km] 5.3 - mm Coal Fine Coal 50 350 1.55 115 At a concentration of 32% a 300 mm pipe is indicated to (-1.5 mm) transport 3 million t/a. This results in a suitable pumping Coarse Coal 32 300 3.6 970 large pipe sizes since the deposit velocity increases roughly velocity of 3.6 m/s. A high pumping velocity is required in Stabilized (- 15 mm) 58 300 1.75 400 with the square root of the pipe diameter [21]. At this veloci- (- 15 mm) Coarse Coal ty the pressure gradient in the 100mm pipe is 1,600 Pa/m Agglomerated 57 300 1.8 115 compared with the pressure gradient for water flowing at Coal this same velocity of 900 Pa/m. Scale-up of such a heterogeneous slurry as this is best achieved by assuming that the

gradient remains constant regardless of pipe size. The wadifference between the slurry and the water pressure 7. Conclusions

ter pressure gradient decreases with pipe size. Calculations Four different options for pumping a particular coal have along these lines result in a pressure gradient of 970 Pa/m been tested in a 100 mm diameter pipe loop. Scale-up of in a 300mm pipe. these results to larger pipe sizes has permitted comparison 5.4 j mm Stabilized Coal Slurry between the different options. It is concluded that only the

fine coal and the agglomerated coal are suitable for long

Fig. 4 shows data for two concentrations, 48% and 58%. distance pumping. The stabilized coarse coal appears suit- In both cases deposition was observed at low velocities. able for medium distances whilst the coarse coal slurry is The scale-up of stabilized coal slurries is not straightforonly suitable for short distance pumping. ward. Thomas [17] showed that below a certain critical These tests were conducted on a raw coal of particularly concentration scale-up by assuming homogeneous laminar high ash content and hence high solids density. The conflow behaviour was not possible. In the present case the clusions could be somewhat different for a low ash coal, lower concentration results could be suspect in this regard. especially as regards the stabilized option. For this reason only the higher concentration data have been scaled up using this approach. At 58% concentration a 300 mm pipe is indicated, giving a References velocity of 1.75 m/s. The pressure gradient would be [1] Rigby, G.R. and Calcott, T.G.: Pipelining Coking 400 Pa/m. Coals; BHP Tech. Bull., Vol. 21 (1977), No. 2, p. 2. 5.5 Agglomerated Slurry [2] George, T.J.: Expanding the Role of Coal Slurries; The agglomerated slurry results of Fig. 5 are scaled up by Coal Min. and Process., Nov., 1982, p. 42. assuming pseudo-homogeneous behaviour as for the fine [3] Orr, L.P. and Thompson, T.L.: Staten Island Coal coal slurry. A 300mm pipe is indicated resulting in a flow Export Terminal; bulk solids handling, Vol. 2(1982) velocity of 1.8 m/s and a pressure gradient of 115 Pa/m. No. 4, p. 661.

[4] Faddick, R.R.: Shiploading Coarse Coal Slurries;

j6. Discussion of Scale-Up Predictions 1982, p. 37. Proc. Hydrotransport 5 Conf., Hannover, F.R. Germ.,

Table 3 shows the scaled-up results for the four options. [5] Rigby, G.R. and Thomas, A.D.: Slurry Handling and The advantages and limitations of each slurry option are now obvious. For this particular coal the only options feasi- Transportation Developments; Proc. Int. Conf. on Bulk ble for long distance pumping are the fine coal and the Materials Storage, Handling and Transportation, I.E. Aust., Newcastle, Aug., 1983, p. 178. agglomerated coal. Both would require pump stations about every 100 km. The agglomerated slurry involves a [6] Derammelaere, R.H., Dina, M.L. and McEwan, smaller size pipe so the capital cost of the pipeline could be P.G.: ETSI Coal Evaluation Plant; Proc. 7th Int. Conf. somewhat less although the cost of the agglomeration on Slurry Transportation, Lake Tahoe, USA, 1982, stage must also be considered. p. 27. The stabilized coal slurry would require pump stations [7] Derammelaere, R.H. and Wasp, E.J.: Dewatering about every 25 km. This probably restricts it to medium of Pipeline Coal; Proc. 9th Int. Conf. on Slurry Transdistances. The increased capital costs and operating costs portation, Rome, Italy, 1984, p. 105. will be offset to some degree by the benefits of the greater [8] Elkes, G.J., Rigby, G.R., Simson, H.A. and particle size especially in an export situation, although Mainwaring, D.E.: Integrated Coal Upgrading and about half of the coal is still very fine and requires expensive Slurry Transport; Proc. 10th Aust. Chem. Eng. Conf., de-watering. Inst. Engrs. Aust., Sydney, 1982, p. 109.

The coarse coal slurry is obviously limited to short distance [9) Rigby, G.R., Jones, C.U. and Mainwaring, D.E.:

pumping. A centrifugal pump is required every half kilo- The Potential for Coal Slurry Transportation and Benemetre or so. The high pumping velocities will cause very ficiation Operations in Electricity Generation; Proc. high pipe wear rates. Inst. Engrs. Aust. Eng. Conf., Newcastle, 1983, p. 92.

667

Page 6

Slurry & freight pipeline technology Volume 8, Number 6, December 1988 solids bulk

[10] Rigby, G.R., Jones, C.U., Mainwaring, D.E. and [16] Brookes, D.A. and Snoek, P.E.: Stabflow-Slurry Thomas, A.D.: Slurry Pipeline Studies on the BHP- Development; Proc. Hydrotransport 10 Conf., Inns- BPA 30 tonne per hour Demonstration Plant; Proc. bruck, Austria, 1986, p. 89. Hydrotransport 8 Conf., Johannesburg, South Africa, p. 181. [17] Thomas, A.D.: Pipelining of Coarse Coal as a Stabil- [11] Rigby, G.R., Yoshinaga, M. and Mainwaring, ised Slurry - another viewpoint; Proc. 4th Int. Tech. Conf. on Slurry Transportation, Las Vegas, USA, 1979; D.E.: Flow Properties and Utilisation Prospects for p. 196. De-oiled Coal Agglomerates Produced by Oil Agglomeration; Proc. 4th Int. Symp. on Agglomeration, Toron- [18] Duckworth, R.A., Pullum, L. and Lockyear, C.F. to, Canada, June, 1985, p. 867. The Hydraulic Transport of Coarse Coal at High Con- [12] Yoshinaga, M. and Rigby, G.R.: Compaction and centrations; Proc. 4th Int. Symp. on Freight Pipelines, Degradation of De-oiled Coal Agglomerates During Atlantic City, USA, 1982. Sea Transportation; Proc. 13th Aust. Chem. Eng. [19] Lockyear, C.F., Pullum, L., Duckworth, R.A. Conf., Inst. Engrs. Aust., Perth, Aug., 1985, p. 69. Littlejohn, M.H. and Lenard, J.A.: Prediction of (13) Thomas, A.D. and Rigby, G.R.: A Study of Some Pressure Gradients for the Transportation of Coarse Factors Affecting the Rheology of Coal-Water Slurries; Transportation Conference, Perth, Oct., 1984, p. 65 Coal in a Fine Coal Carrier; Proc. Inst. Engrs. Aust. Proc. 13th Aust. Chem. Eng. Conf., Inst. Engrs. Aust., Perth, Aug., 1985, p. 443. [20] Duckworth, R.A., Pullum, L. and Lockyear, C.F.: [14] Ercolani, D.: Production Plants and Pipeline Sys- The Pipeline Transport of Coarse Materials in a Nontems for Snamprogetti's Coal Water Slurries. Recent Newtonian Carrier Fluid; Proc. Hydrotransport 10 Experience and Current Projects in Italy and USSR; Conf., Innsbruck, Austria, 1986, p. 69. Proc. Hydrotransport 10 Conf., Innsbruck, Austria, [21] Thomas, A.D.: Scale-Up Methods for Pipeline Trans- 1986, p. 19. port of Slurries; Int. Jnl. of Mineral Processing, Vol. 3 [15] Bhattacharyya, A. and Imrie, l.: Development of (1976) No. 3, pp. 51-69. the ASEA Mineral Slurry Transport System for Coarse [22] Thomas, A.D.: A Rational Design Philosophy for Long Coal; Proc. Hydrotransport 10 Conf., Innsbruck, Distance Slurry Pipelines; Chemical Engineering in Austria, 1986, p. 63. Australia, Inst. of Engrs. Aust., 1977, pp. 22-33.

The Compact Line

for palletless packages -

a new generation!

The space saving and most economic new compact line for palletless shrinkwrapped packages of bags, cartons, cases and other stackable goods concentrates all features for placing, shrinking and profiling of the film hoods in one single working station from only one film reel

möllers

Maschinenfabrik Möllers GmbH u. Co, D-4720 Beckum/F.R. Germany, P.O. Box 1764, Phone (02521) 88-0, Telex 89423, Fax (0 25 21) 88-100 Subsidiaries in Brazil, USA, Great Britain und France