Ol, STORAGE TANK FOUNDATIONS ‘Too often the importance of foun~ dations is underestimated. Since the flat bouemed oil tank is one of the simplest types of structure and has suiticient ilexibility to accommodate appreciable changes in shape, the Foundation on which it rests is fre- quently given Hitle consideration, With “the advent of larger and higher tanks, the esfects of poor foun- mis were multiplied, ‘The larger ‘not only imposes a heavier load and hence causes greater settlements, tur the distortion caused by any given settlement is, in some measure, pro- portional to the size of the tank. Ti auidition, the floating roof user has become inereasingly conscious of the importance of a shoe-to-shell fit that is xs nearly perfect as possible, Temiust be obvious that perfection can best be atlained if the tank shell is baile civewar and remains cireular within reasonable limits. TT the grade is not level in the be- ginning, or if 1 later settles unevenly, the tanks will inevitably have a dist torted shell. Often the tank builder is blamed for a poor shell that should properly be charged to a poor founda- tion, Tn order to obtain. good tanks, good foundations must be provided Proper grade preparation can also have an important bearing on bottom corrosion. Tans erected on poorly Grained grades, directly contacting cortosive sols or on heterogeneous nistures of different types of soils are all subject to electrolytic atack on the bottom side. By leving E. Bobera. Chiet Engineer There is no rule of thumb that can bbe applied to all situations, Each case ‘must be evaluated om the basis of con- ditions as they exist. Local conditions vary so widely that it would be im- possible to anticipate them all. An attempt has been made in the material that follows to deal with the more common aspects of tank foundation desiga. In Selecting the proper type of foundation, the bearing power of the soil is the primary factor. Where no previous experieriee in de same area is available, soil borings to determine ‘existing conditions are usually cheap ingurance against future trouble, We have seen a number of instances where tank sites were judged solely from surface conditions only to have the empty tanks settle so seriously during construction that the water test could not be performed until the foundation was rectified. With the tanks already erected, this could only be accomplished at great expense. We know of one instance where a tank settled under water test to stich aan extent that it collapsed. While these are extremes, they serve ta illustrate the importance of first knowing the nature of the foun dation dase. Knowledge of geological formation or experience with other heavy structures in the same vicini will often suffice, but if such knowl ‘edge is absent, Soil borings are the safest means of investigation. There are many firms over the country well skilled in the art of making such bor- ings. Sofl beating tets are now yen erally conceded to have litle vale Giher than to evaluate conditions at the tested level, ‘The depth to whieh fn applied Toad will produce sigoii- tant stresses isin some measure pro- Dortional to the size of the loaded Brea. A large tank fully loaded will probe out wealmesses never revealed bya test on 1 or 2 square fect Frequently. the resule of borings will inicate the desirability of limite ing tank height rather than the cx pense of costly foundations. ‘Assuming that bearing conditions have been determined to be adequate, the simplest form of foundation is 2 Sand pad laid directly on the earth ‘All loam or organie materia] should tbe removed and replaced with suit able material, well compacted. Often Sa satisfactory fill material is available atthe site. if not, banks run gravel is excellent and is readily ‘compacted, “The grade for the tank sl preferably be elevated slightly sicve Ihe surrounding terrain. to Grainage, ‘Sullicient bers shou provided. to prevent ws Wweathering under te tank djerm width should be at Te Weathering can be minimize erm is subsequently protected. with tray tock, gravel, of at asphaltic ashi Tt is Customary to provide a crown of about 1 in, for each 10 i, of ra- dius. On large tanks, the erown is sometimes Timited to 6 in ‘The wand pad should be at least Jore placing sand cushion, sadreplaced with well cc stone "aN ano uss TYPICAL TANK GRADE -ojige ox ctor orale maar shouldbe semeval from te pe oy ind hi “Frequently the entire area is stripped and the strippings used in cons tig Oe Reewall’ Bers shosla’ be. protected against weather wih we ‘rock, gravel or asphalt,4 in, deep. ‘The sand should be clean seul” iree from corrosive elements, ‘are should he taken to exclude clay ‘or Tumps of earth from coming into contact with the bottom. “Frequently the difference in potential between two types of earth will set up an elec- Urolytie eell with ressliant pitting. ometimes erushed rock is substi= tuted for sand, but sand is easier to grade and usually more available, While there is some ditference of ‘opinion, it is considered desirable to olla sand grade, Its not satisfactory to merely pour oil on the surtace be- dause id will not penetrate or mix with the sand, Subseyuent laying of the bottom wilt be dificult and weld- ing is likely 0 start a fire. Oil and sand should be thoroughly mixed either in_a conerete mixer or by windrowing with a grader or by rand. Only’ sufficient oi should be used to thoroughly wet che sand, E> cess free oil will make subsequent working on the grade both diftcalt and hazarious. 1 the snd cushion is placed on top crushed rock iil, the rock should carefully graded from coarse at hhottom to line at the top. LE this is, ne, the sand will percolate down sirnagh the voids in the e An excellent base ca iaied Dy substituting about 134 in. Of asphalt “oad paving mix for the sand cushion. ‘This material 3 availe able from :eady mix plants in many sections of tie country, ‘There is one precaution that must’ be carefully watched. After the material has set up, it is difficult #f not impossible for the tank buicer to correct inaccu- racies by taking down the higi and filing in the low spots. Tt is, there fore, most important that a paved tank grade be levelled with extreme accuracy under the shell. Drainage is important both from the standpoint of soil stability and bottom corrosion. Good drainag’ should be provided not only under the tank itselz, but she general area should preierably be well drained, Where the terrain does not afford natural drainage, proper ditehing around a group of tanks may help to correct Ue vel Where suitable bearing soit is not avaliable at the surface, but is avail- able a reasonable distance below the ‘ing wall foundation is The purpose of the ring is, the sail and prevent lateral height of sell beight of ring wall Above grade) “8 T (thickness of wall) = — “Teal bap tens = PR = Required area of hoop steel Use 24 bars Win, diam, 4 = weight of stored product aon steel weight on wall = 1520 Ibs. per ft. il 5 ft (assumed 4 i. frost tine and 1 45 tbs. per eu 24 ein 5x03 ((625 x 48) + "6g75 = 75 = 3669001 ‘00 . “nag = 165 NOTES Ring wall to be founded below frost line. Note that Weight of stored pradct wat use i determin, Nieves weight of Water was used i determining reintoreing ‘Sect. Tes hchiewe that settlement 4 fae tor of Hie gs el os a than host cases the Tong’ term loading iy more Signidcant chan the temporary ‘water Toad tudor test, Or the ther fan the hoop ten= Son de to water tent oad must be taken Into secount to avoid cracking of wall ‘Where the characteristics of the confine soil are known, the designer shod use the correct ratio hlween vertical and horveontal Bressures rather than the safe limit of 03 [sed above. For the sbalow wall ilusteated, the hoop tension was assumed uniform from top 10 Bottom of walt For high walls, the increase in lateral pressure with depth would reasize loser steel spacing at the bottom Wan at the tor movement. ‘The ving wall is founded jn the firm stratuny and confines the weaker materials, Totally inadequate material should be removed and re- placed with well compacted fil. ‘Many tank owners wse ring wail consiraetion as standard even” when soil conditions do not indicate its use. There area number of advantages in this practice that may well, compe sate for the added cost, ‘The imei- dental advantages of the ring wall are neat appearance, an excellent founds tion for the tank shell, the elimination of washing and weathering of be and the exclusion of surface water Funning into the grade. ‘We recommend that ring walls be so proportioned that the unit soil beating at the level of the bottom of the wall is the same under the con- crete as under the confined soil. This ‘an be accomplished and will encour age uniform settlement of the foun- dation as a whole ‘Assuming the tank shill to be cen- tered on the walt, and earth to weigh 100 Ibs. and concrete 140 Ibs. per ceubie foot, the thickness of the con- rete wall’ may be determined from the following formula Thickness of wail ZW aH - 80h Where W = nce of mata nae an rot Bega te ae wal BREET inches = welghe of stored product ib pero. sipoe te Seu The ing wl es hae 3h On tanks of relatively small diam- eter, the formula will result in ring walls less than 8 in. thick, in shiek ease the 8 in, minimum ‘should. be used as the closest approach to the desired ideal Where a ring wall js used, it shoul! be reinforced circumferentially to ie- sist the hoop stress sesulting. from lateral pressure of the confined ear Because soil conditions are rarely known in advance, it is our practi to design such walls on the basis a lateral pressure equal t0 03 of 1 combined liquid and earth vertical pressure. For shallow rings the ver- tical load contributed by the earth is small, but on deep walls it can be come’ important. If there are openings in the wal the reinforcing’ must be carried around such openings to preserve the continuity of the hoop action, Non inal vertical steel is normally pro- Vided primarily for convenience placing hoop steel. Through the years there has been considerable discussion of the merits of placing the tank shell on the con ‘rete versus placing the tank entirely within the ring wall, For many years we contended that the tank should not rest on the wall TU was feared that if settlement was greater within the wall than uoder the wall, failure of the bottom could occur. ‘On a carelessly prepared grad this could happen and has happened It has been our observation, however,