TECHNICAL FORUM ey Alkali-to-sulphate ratio, build-ups and rings See CCS h eR Ue ececu remain crea ae Rec cey Ce PUTA conference in Singapore on a variety of topics. One presentation about reducing cyclone etchant Con emut ie lelmeke oN miata cmet et sycae aerate Ca tthe Cemtech Asia 2017 conference, Siogtried Ande of Sick AG, Germany, presented the paper: "Reducing cyclone blockages using hot wet process gas analysis in which he ported ona six sigma projectundertaken by Gimsa {Turkos) nto the causes of preheater ‘ycione blockages. ima found no correlation between cycione blockages and the variation inthe inputs othe kiln, However, the company installed anew kin inlet probe and gas analyse rom Sick to moritor SO, at then inet It foune that he SO, at then inlet ses to>10,000ppm then the cyclone blockage likelihood rose to >80 per cent. The gas analyser gives the coment producer advance warning of theintensfcation ofthe sulphur eycle between the rotary sectionofthekiln and the preheater. The process can then be adjusted to reduce the intensity ofthe sulphureyce, the SO, atthe kin inlet and thelikelihood ofa preheater blockage Role of oxygen Themain causes of this increased SO, at klninlet were the main burner flame shape and whether burning conditions in the kiln were oxidising or reducing. This isno surprise as tiswel understood that the stability of calcium sulphate atthe temperatures in a cement kilnis closely linked withthe avaitabity of exygen in the ln atmosphere. The combination of SO, in thekiln atmosphere with oxygen andthe lime present in the hot feed in the lower stages of the preheater isa reversible reaction C280, ++ CaO + SO/t + 10,4 If theresa shortage of oxygen at any pointin the kiln where the temperature is 2900°Cthen the reaction as written is drvento the products breaking down calcium sulphate (CaSO), and liberating free lime (CaO), SO, and oxygen, 0, to rebalance the oxygen equilibrium, This breakdown of calcium sulphate is one of the reasons why heawy sulphurcycles ‘and reducing burning conditions can be associated with the kiln producing high free lime clinker. Ifthe reducing conditions are at the main burner of the kiln then the lime liberated from the breakdown of calcium sulphate does not have time to recombine with beite,C,S,to formalite, CS. Alkali-to-sulphate molar ratio In the July issue oF ICR, the moderator also noted the article: “Preventing bulld- upsand rings" by Tom Lowes and Joana Bretz de Souza ~the same topic addressed by Mr Andra Singapore. The authors make the point thatthe alkal-to-sulphate molar ratio isthe key to understanding the susceptibility ofa particular kin to the formation of build-up inthe preheater and rings in the calcining or upper transition zone of the kil. The equation forthe alkalito-sulphate molar ratios: (96K, 0/94 + aeNa,0/62-%cI" 73) (9480,/80) where the factors relate tothe molecular and atomic weights ofthe volatile species. To avoid hard preheater build-up and ring issues ths alkal-to-sulphate rmolarratio should bet. In that case the sulphate will be bound as the relatively thermaly-stable Na,SO, and K,S0,, Which tovolatilse need temperatures of >1600°C, well above those normally found ina cementkiln. Bound withthe alkalis, ‘thesulphate will therefore pass out of the kiln inthe clinker and there will be no calcium sulphate present to break down under reducing conditions the kiln So the higher the alkalis with respect ‘to sulphate, the better for avoiding hard build-up and ring issues, Thatis not to say that there will be no build-up with alkalis wellin excess of sulphate, butalkal-based build-ups are soft and easy to clean, while calcium sulphate build-ups are hard and therefore difficult to clean, The importance of chloride The article's authorsalso made the point that the alkal-to sulphate molar rato inthe hot meal nthe klnis the ctical balance and thatthe chloride corection must be included. The alkalis wil preferentially combine with chloride rather than sulphate in the hot eal. Therefore, the presence of chloride takes up alkalis, eaving less alkali available to combine wth sulphate and as a result leaves more sulphate availabe to form calcium sulphates. Similarly, the moderator was advising ‘a cement company in the Americas on 8 ‘ing issue nits kin and asked for data on the hot meal chemistry. This was provided forallalisand sulphate but not chloride, which was not measured. That rendered the exercise a waste oftime as chlorides are extremely volatileina cement kin and will build to high levelsin the hot meal, even with very smallinputs of chloride to the in the chloride correction isnotFigure2: concrete block re the North Sea ater several yea ‘aggregate which, in the case of underwater constructions, was preferably also volcanic ‘material with pozzolanic properties. The concrete is described by Pliny the Elder as becoming a single stone mass as soon as itisin contact with thesea, This may have been a little exaggerated, but the modern reconstruction of a Roman concrete block inthe sea at Brindisi reported in ‘Bullding, for Eternity’ reports that the upper layers. ‘ofthe test block could be walked on the day after placement. How doesit differ from modern concrete? ‘The basis of both modern cement use and the Roman cementitious mixes s the calcination of calcium carbonate toproduce calcium oxide and then its reaction with water and sitca to produce calelum slate hydrates. tis necessary for the silica tobe in aform that will dissolve in the alkaline water produced by the dissolution ofthe lime, Portland cement uses an intermediate stage of combining the lime and silica at high temperatures in cement kiln to produce calcium slcates, wiih then dissolve in waterat ambient ‘temperature and reprecipitate as calcium ‘slicate hydrate A large number of modern concretes also supplement the Portland cement with ether fly as from coal- fired powerstations or with granulated blastfurnace sag (68S) from steel plants. Fly ashis pozzolanic in asimitar manner to the Romans’ volcanic ashes, while GBSis latently hydraulic, meaning itwill react with water in an alkaline environment to produce similar reaction products to Portland cement. So while Portland cement needs to be heated to approximately 1450°C to produce catelum silicates, Roman coment used a material inwhich an alumino-sicate component vered from the bed of was already soluble inalkaline water, ‘namely ash which trad erupted from a volcano and rapidly cooled in the ar. An added advantage of thiswas that the mosteffective pozzolanas were ‘extracted from the ground asa powdered material eliminating theneed for milling, significant difference between Roman and modern concretesis in the pore structure and the relationship to seawater. Because the Roman concretes did not contain steel reinforcoment the presence of chloride was of lesser significance in terms of durability than Inmodern structures. tis likely that the ‘mortarto be mixed with coarse aggregate was mixed using seawater: In the model described by Jackson etal the dissolved species in the seawater formed an integral part ofthe continuing hydration cycle ‘whereas the main thrust of producing modem steel reinforced concrete for installation in the sea isto keep the seawater out. ‘The maritime environment Itwas discussed above that, apart from the ingress of chloride and CO, which depassivitate steel reinforcement and causerusting and subsequent degradation of the concrete, the main damage to ‘concrete in the sea is found either in the intertidal zone or the splash zone above the high tide mark (Figure 1. “Building for Eternity’ lists 93 examples ‘of Roman concrete maritime structures. ‘Apartfrom the remains of a Roman fish pondin Portugal, all the sites reported inarewithin the Mediterranean. The ‘Mediterranean Seaiis a sea with very limited tidal range, mostly ofa few centimetres, so the extremes ofthe twice daily wetting and drying cycles which would affect coastal concretein most of the world, have not been experienced by ‘the Roman concrete samples examined and recorded, However, the Mediterranean Sea does experience sovere storms which ‘would produce the splash zone that causes damage to modern concretes. During the 350 or so yearsin which Britain was ruled from Rome a large ‘numberof land-based buildings were constructed, many of which survivein ‘some form today, butthere appear to be no known remains of structures built under the sea during ths time. It seems unlikely that there was no effortput into constructing quays or breakwaters ‘and possible that these were built but had not survived as wellas those inthe editerranean, I they were constructed, possibilty for theirnon survival could be the shortage of volcanic tuffto use as coarse aggregate in Britain. Thepozzolana in Romano-British buildings was frequently supplied by grinding up fred clay tls, but these would not have contained the zeolites and feldspars referred to by Jackson as providing he authigenic mineral cycling. Asecond pessibility would be that the environmantin Britain, where the conerete would be subjected tothe vagaries ofthe southwesterly winds, the constant wetting and drying cycles above low tide and the large temperature range was too much for ths desig. inany eventitwould be very instructive, ifthe Severn tidal barrage is tobe built include a Roman concrete design to try to match the concrete specification. The construction would need tobe able to translate the Roman methodology, mixing and placing by hand with the ai of avery large slave labour force, into something using modern methods to produce the desired pore structure. The Romans may have missed them, butthere are volcanic tffs in Wales which ‘may be suitable, although many are now within national parks and therefore, unlikely tobeavallabe, REFERENCES. "BRANDON, CJ, HOHLFELDER, RL, JACKSON, Mt AND OLESON, JP (2014) Building for eternity: the bistory and echnalogy of Roman concrete ‘engineeringinthe se Oxors,UK: Oxbow Books. "HAREISSON, AM (2010) "Fy shin cement in (CR December, 57 > HARRISSON, AM, WINTER, NB AND TAYLOR, HF. (1986) 'An examination of some pure and composite Porlandcement pastes sing canning electron microscopy wth Xray analytical pati’ in: sth ccc proceedings lola Rie de anata, pli “JACKSON, MD, MULCAIY, SR CHEN LY LU, Q,CAPPELLETTI,® AND WENK, H-% (2017) ‘Philipste and Altobermorte mineral cements produced through low-temperature water Fock reactions Roman marine concrete’ In ‘American Merton, 102 (7), L435. ® BUENFELD, NR ard NEWMAN, J (1984) “The pemesbity of corereteina marine cemironment in Magazine of Concrete Research, 36 (127, p78.