Recommendations of Workshop on Revision of I.S.

Codes on Liquid Retaining Structures

November 10 and 11, 2005 DEPARTMENT OF CIVIL ENGINEERING INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Comments on the recommendations or any feedback may please be sent to the following persons: Dr. Ashok K. Jain Professor of Civil Engineering Indian Institute of Technology Roorkee e-mail : ashokfce@iitr.ernet.in yahoo group: http://groups.yahoo.com/group/structure_iitr Phone / Voice mail: 01332-285436 or Dr. Vipul Prakash Asst. Professor of Civil Engineering Indian Institute of Technology Roorkee e-mail : vipulfeq@iitr.ernet.in Phone / Voice mail: 01332-285538

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Introduction Each year hundreds of all kinds of water tanks and other liquid retaining structures are being built in different parts of the country. These tanks and reservoirs have faced many storms and earthquakes. During the past four decades there have been considerable developments in materials, design philosophies, construction techniques as well as attitudes of all the stake holders. Therefore, now the endeavour of the engineers is to produce stronger, durable, economical and aesthetically appealing liquid retaining structures. There have been a few instances of failure of tanks and reservoirs. The reasons are many. There is a need to understand the causes of failures and ways to prevent them in future without punishing good designers, field engineers and contractors. The objective of the workshop is to bring planners, designers, contractors, and various state government agencies associated with such structures on a common platform and help evolve better codal specifications. Participation The Workshop was represented by the Chief engineer, Suprintending engineers and Executive engineers from U.P. Jal Nigam, Uttranchal Peyjal Nigam, Punjab Water Works and Haryana Water Works. Besides these very experienced officers, Prof. Anand Prakash and Prof. S. K. Agarwal, former professors of the IIT Roorkee, who have designed thousands of LRS over the last 40 years in the various states of Northern India also participated very actively. Performance of LRS One full session was devoted to the performance of various types of LRS during gravity and lateral loading during the past 50 years or so. The following was the general consensus: (1) It was pointed out that in Haryana most of the tanks were shaft type. In Punjab, most of the tanks were on beam-column frame staging. UP and Uttaranchal had mixed type of tanks. (2) No LRS designed in accordance with IS:3370-1965, properly detailed, properly compacted, cured and constructed as per good construction practices has failed under gravity and normal lateral loads. This was the unanimous view of all the engineers from Uttar Pradesh, Uttranchal, Haryana and Punjab. (3) Detailed survey reports of LRS during the Jabalpur earthquake 1997, Bhuj earthquake 2001 and the recent October 2005 POK earthquake showed that no LRS ground supported, overhead tank on beam-column frame staging or on shaft staging has

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

failed due to an earthquake. The tanks in the Kashmir region near the LOC were designed and constructed on shaft staging by MES. All the tanks escaped unhurt. (4) It was further pointed out that a few tanks have failed during the first filing itself. It was attributed to poor concreting excess w/c ratio, inadequate compaction and curing. During earthquakes LRS with shaft staging developed horizontal cracks at the level of lifts, that is at construction joints. (5) A few LRS with shaft staging failed due to the storage of bleaching powder, used for chlorination of water, inside the shaft staging. (6) A report of the World Bank on the performance of structures during the Bhuj 2001 earthquake concludes as follows: Elevated Water Tanks. These stand-alone structures consist of both the traditional RCC supporting frame type of tanks, as well as the more modern inverted pendulum type structures. In most cases these appear to have been designed for lateral forces and have survived without significant distress, even near the epicentral region. (7) A report on the Jabalpur earthquake of 1997 at the NICEE web site (IIT-K) concludes as follows: None of the OHW tanks supported on MR frames were damaged even though most may not have been designed for the seismic force. This is in line with what was observed in Latur 1993 Earthquake IS:3370-part 1 The draft code was discussed clause by clause. The recommendations are as follows: (1) It was pointed out that the climate in UK is quite different compared to that in India. Clause 1.2 states that this British Standard (BS:8007-1987) applies particularly to the UK conditions, and although the principles are applicable to design in other parts of the world, the designer should take account of the local conditions, particularly variations in climate and the possibility of earthquakes which have not been considered for UK conditions. Considerations have been to the storage of liquids at ambient temperatures or at temperatures up to approximately 35oC such as are found in swimming pools and industrial structures. Clause 2.7.1 of BS 8007-1987 says that, For a correctly designed structure and good quality materials and workmanship, the design life of the structure should be between 40 years and 60 years. Some components of the structure (such as jointing materials) have a shorter life than the structural concrete and may require renewal during the life of the structure.

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

The existing specifications in IS:3370-1965 are already catering to the design life of over 50 years in India. (2) In view of the performance of the LRS during the past 50 years, designed in accordance with IS:3370-1965, there appears to be absolutely no justification for increasing the minimum grade of concrete from M20 to M30. All the participants felt very strongly that M20 grade of concrete must be retained as increasing the grade to M30 will only lead to theft of cement in practice. It was pointed out that it would not be possible to produce M30 concrete at site for overhead tanks of small capacity. This would require revision of Table 1 in Part 1 and Tables 1, 2 and 3 in Part 2. In Table 1 of Part 1, for pre-stressed concrete tanks the minimum grade of concrete needs to be M 35, in line with IS:1343. It was pointed out that the minimum reinforcement is governed by Annexure B as follows: B-1.2 Minimum Reinforcement To be effective in distributing cracking the amount of reinforcement provided needs to be at least as great as that given by the formula : crit = fct / fy where crit = the critical steel ratio, that is, the minimum ratio, of steel area to the gross area of the whole concrete section, required to distribute the cracking. fct = the direct tensile strength of the immature concrete (usually taken at the age of 3 days as 1.15 N/mm2 (MPa) for M 25 and 1.3 N/mm2 (MPa) for M 30 grade); fy = the characteristic strength of the reinforcement. Therefore, if the minimum grade of concrete is increased then it has the undesirable effect of increasing the minimum percentage of steel required. (3) There appears to be absolutely no justification for increasing the minimum cover from 25 mm to 45 mm. A moderate exposure condition as specified in IS:456-2000 appears to be quite justified and must be retained. This will entail review of Table 1 of Part 1, specifying minimum cement content, water-cement ratio and minimum grade of concrete. In Table 1 the minimum cement content should be 360 kg/cu.m. for M20 concrete and the maximum cement content should also be specified owing to the cracking due to shrinkage and heat of hydration. (1)

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Increased shrinkage cracking was reported owing to use of higher grades of cement, besides their low shelf life. 33 grade of cement was to be preferred. The tank body is designed on a no-crack basis with low values of permissible stresses in concrete and steel, and therefore the cover requirements given in IS:456-2000 wherein design is based on Limit State of Collapse are much on the higher side. It was suggested that IS:3370 should specify the cover requirements for different exposure conditions for designs on no-crack basis. It was suggested that the exposure conditions defined in IS:456-2000 be retained, but the cover requirements corresponding to one-grade lower exposure be used for designs on no-crack basis. It was also suggested that Portland ground granulated blast-furnace slag cement with slag < 60% be recommended for the construction of the tank, as it offers the maximum resistance to chloride penetration a common phenomenon owing to the necessity of chlorination of potable water. (4) It was pointed out that since the joints are sources of leakage, movable joints were never used in LRS in these states. Only construction joints and expansion joints in very large tanks were used. Part 1 of the Code deals only with concrete and joints. It may be merged with part 2 of the Code. Only the treatment of construction joints and expansion joints may be retained in the main body of the Code and other types of joints may be placed in the annexure, if required. (5) Few clauses need to be added for lightening conductor for OHT, for ventilation particularly in GWR, for lining of inner (liquid retaining side) with some sealing chemical/materials, etc. IS:3370-part 2 The draft code was discussed clause by clause. The recommendations are as follows: (1) There is a need to specify minimum thicknesses of various members of an overhead LRS in the Code for ensuring proper placement of reinforcement. The following thicknesses are proposed: Top dome 100 mm Top slab 120 mm Vertical wall 120 mm at top and 150 mm at bottom Bottom dome/ bottom slab 150 mm Conical/slant wall 150 mm It was pointed out that higher thickness for top dome/slab is required because of high chlorine concentration.

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

(2) The tension in steel on face away from the water face should be retained as 190 MPa in HSD bars. It was also resolved that TMT bars be also permitted, but with same permissible stress values as for HSD bars. (3) The minimum steel should be based on the formula given in Annexure B of the surface zone with further reduction with increase in thickness as per the existing specifications.. Up to 150 mm thickness the %steel may be calculated as per formula. As the thickness increases linearly reduce it to two-thirds of this value for 500 mm thickness. Linkage of minimum %steel with diameter (15 m) be deleted. Surface zones be defined to be of 150 mm thick only. Any increase in minimum %steel requirements will cause heavy congestion of steel at the middle ring beam joint and bottom ring beam joint. It would also cause congestion problems in the design of conical tanks at the vertex of cone. It becomes very difficult to pour concrete at these joints even when the steel is 0.24%. It was pointed out the minimum percentage of steel is governed by the formula given in Annexure of IS:3370 (Part 2). According to this formula, higher percentage of steel is required for higher grades of concrete. Therefore, it is important not to increase the minimum grade of concrete for LRS. (4) It was resolved that Clause 4.4.3.1 for limiting the crack widths be modified. To control crack widths it was important to provide smaller diameter bars at closer spacing. It was recommended that if the maximum spacing of bars is limited to 150 mm, then no crack width calculations will be necessary. It was also recommended that the minimum diameter of steel bars be 8 mm. (5) Regarding clause 5.2, it was resolved that the screed layer be as specified earlier, i.e., minimum M 10 concrete and minimum thickness of 80mm or 100 mm. (6) The limit state design of the container in the present form appears to be mostly of academic interest. (7) Annexure A needs to include IS 875, IS 383, IS 2309 (lightening conductor). (8) The foreword of IS:456-2000 on page 2 says that In this code it has been assumed that the design of plain and reinforced cement concrete work is entrusted to a qualified engineer and that the execution of cement concrete work is carried out under the direction of a qualified and experienced supervisor. The same statement should be included in IS:3370-part 2 in place of the existing statement. IS:11682 Although this code is not under revision, there is a need to include an equation for computing the stiffness of beam-column staging.

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Prof. S.K. Agarwal suggested the following equation: K= 12 Ec I col N c

h
p =1

Np

for bracings at unequal intervals

3 p

= Where,

12 Ec I col N c for bracings at equal intervals N p h3

Ec = modulus of concrete I col = moment of inertia of column = D4/64 N c = Number of columns in staging

N p = Number of panels in staging along the height = N b + 1 N b = Number of bracing levels in staging along the height hp = h = c/c height between bracing levels.

= 1+

2 Nc sin 2 Nc

Stiffness obtained from this expression is about 8 to 12 % more that obtained from 3D analysis of staging. IS:1893-part 2 The draft code was discussed clause by clause. The recommendations are as follows: (1) There is a drastic increase in seismic force without any logical justification. Even the recent October 2005 POK earthquake M 7.8 showed that not even a single OHT has failed even near the LOC of India and Pakistan. Therefore, the R factors should be de-linked with buildings and appropriately increased so that the existing force levels are maintained. The earthquake force calculated from IS:1893-part 2 should be calibrated with that obtained from IS:1893-1984. For this calibration, the Importance factor for water tanks needs to be higher than 1.0, but may be less than 1.5. (2) The participants expressed serious concern that in case the seismic forces are increased as proposed, what will be the fate of the existing LRS? How will these LRS be strengthened and retrofitted? (3) Based on the above two recommendation, I = 1.2 and R = 4.0 is proposed for water retaining tanks. For Zone V, the figure below shows the comparison for design horizontal seismic coefficient for hard soils as per IS 1893:1984 (using I = 1.5) and spectra as per IS 1893 (Part 1):2002 for I = 1.2 and R = 4.0.

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Comparison of IS 1893:1984 (I=1.5) with proposed I=1.2 and R=4 for Zone V IS 1893:1984 0.16 Design Horizontal Seismic Coefficient 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 0 0.5 1 1.5 Time Period (seconds) 2 2.5 3 Hard Soils Medium Soils Soft Soils

For Zone IV, the figure below shows the comparison for design horizontal seismic coefficient for hard soils as per IS 1893:1984 (using I = 1.5) and spectra as per IS 1893 (Part 1):2002 for I = 1.2 and R = 4.0.
Comparison of IS 1893:1984 (I=1.5) with proposed I=1.2 and R=4 for Zone IV
IS 1893:1984 0.1 Design Horizontal Seismic Coefficient 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 0 0.5 1 1.5 Time Period (seconds) 2 2.5 3 Hard Soils Medium Soils Soft Soils

For Zone III, the figure below shows the comparison for design horizontal seismic coefficient for hard soils as per IS 1893:1984 (using I = 1.5) and spectra as per IS 1893 (Part 1):2002 for I = 1.2 and R = 4.0. 9

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Comparison of IS 1893:1984 (I=1.5) with proposed I=1.2 and R=4 for Zone III
IS 1893:1984 0.07 Design Horizontal Seismic Coefficient 0.06 0.05 0.04 0.03 0.02 0.01 0 0 0.5 1 1.5 Time Period (seconds) 2 2.5 3 Hard Soils Medium Soils Soft Soils

The value of R = 4.0 is proposed for staging with OMR frames, implying no ductility, i.e., 1.0 . Designing a frame/shaft staging for high ductility and lower seismic coefficient is not warranted, because it implies greater tolerance for damage, which is not desirable for Overhead Water Tanks. This value of R is, therefore, for all types of staging masonry shaft, RC shaft, frame, steel, etc. Table Importance Factor, I
Type of tank Tanks used for storing hazardous materials, inflammable or poisonous gases or liquids Tanks used for storing non-volatile material, low inflammable petrochemicals, etc., or for storing water for emergency services such as fire fighting services, hospitals, etc. Tanks used primarily for storing water for drinking, irrigation, etc. All other tanks with no risk to life and with negligible consequences to environment, society and economy. Importance factor I 1.75 1.50

1.20 1.0

In IS 1893:1984 for tanks used for storing inflammable or poisonous gases or liquids, the importance factor I = 2.0. In draft IS 1893 (Part 2), I = 1.75 has been proposed for such tanks and with R = 4, the design horizontal seismic coefficient compares as shown in figure below.

10

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Comparison of IS 1893:1984 (I=2.0) with proposed I=1.75 and R=4 for Zone V for Tanks storing Hazardous Materials
IS 1893:1984 0.25 Design Horizontal Seismic Coefficient 0.2 0.15 0.1 0.05 0 0 0.5 1 1.5 Time Period (seconds) 2 2.5 3 Hard Soils Medium Soils Soft Soils

(4) Concern was expressed that the procedure for seismic analysis was being needlessly being complicated for considering hydrodynamic effects, even though their contribution to seismic forces was less than 15% to 25% in most cases. The imprecision in quantifying the seismic forces is far higher, and does not warrant such sophisticated analyses. A 33% increase in stresses for the design of container will easily absorb this hydrodynamic force. Indeed clause 5.2.7.1 of IS 1893:1984 left the consideration of hydrodynamic effects at the discretion of the designer in the following words:
When a tank containing fluid vibrates the fluid exerts impulsive and convective pressures on the tank. The convective pressures during earthquakes are considerably less in magnitude as compared to impulsive pressures and its effect is a sloshing of the water surface. For the purpose of design only the impulsive pressure may be considered.
Concluding remarks

The country has tremendous experience of design, construction and maintenance of hundreds of thousands of LRS of various shape, size and arrangements. There appears to be no justification for borrowing the specifications from UK, USA or other countries. Any upward revision in the specifications should be based on logic, and Indian climate and experience. An unnecessary increase in the cost of construction will save precious national resources.

11

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Appendix A Comments from Prof. S. K. Agarwal, Former Professor, IIT Roorkee

Before commenting on the codes, it is pertinent to emphasize that the discussion given below relates to water storage tanks & reservoirs. In our country, normally the water is stored in overhead tanks or in semi/fully sunk ground reservoirs. Shapes of containers for overhead tanks are either circular or of Intze type. These shapes are obtained by assembling a number of shells of revolution and/or circular plates. These containers are supported on staging, which may be a framed staging or a shaft type staging. These containers & their staging are normally constructed in RCC. Similarly semi/fully sunk ground reservoirs may be either rectangular or circular in plan. Normally these reservoirs are directly supported on ground and constructed in RCC. Thousands of such overhead tanks and ground reservoirs have been constructed in India. I had been associated with the structural design of a very large number of overhead tanks (of capacities varying from 25 kL to 4000 kL) and ground reservoirs (of capacities varying from 50 kL to 15000 kL). I have also examined 7 overhead tanks, which failed during testing and about 30 overhead tanks, which were under use but severely distressed. Out of these about 50 % tanks have been repaired and are working satisfactorily since last 8 to 10 years. Remaining tanks were found to be beyond economical repairs. Similarly, about 6 underground distressed reservoirs have been repaired, restored and are in use. It is pertinent to state that none of the over-head tanks or reservoirs has failed due to earthquake in about last 40 years for which details are available. All the overhead tanks failed during testing were supported on shaft type staging. Failures had been mainly due to faulty design, poor quality of construction or due to excessive foundation settlement. Some shaft type overhead tanks failed due to corrosion because of poor maintenance or improper use of space in the shaft. Many of the distressed overhead tanks examined were found to contain about 20 to 30 times the required reinforcement but very poor quality of concrete, which led to the corrosion of reinforcement mainly in columns and braces. In a few cases, some cracks were seen in the container portion also. Number of tanks failed or under distress forms a very small (about 1 to 2) percentages of the total number of tanks constructed. All these tanks and reservoirs have been designed on the basis of existing codal provisions. As such, there is necessity to examine the proposed codal provisions to bridge the gaps if any and to update the codal provisions to eliminate (or minimise) such failures and distressed cases.

12

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

It may be noted that the codal provisions may be planned in such a manner that the designs evolved fulfil the essential requirements of a good engineering design, i.e., the designs are efficient, economical and are easy to construct. Engineers, contractors and administrators should not be forced to waste the scarce resources of the country due to the provisions of codal provisions.
1. IS: 3370 (Part 1)

The proposed draft lists one major deviation from the existing code by mentioning in Para 4 (Exposure Condition) as severe as per IS 456: 2000. According to IS 456: 2000, this means that the minimum grade of concrete mix for container portion shall be M 30 in place of M 20 and the minimum cover to reinforcement will become 45 mm. IS 456: 2000 states Exposure conditions in Table 3 as below: Moderate Concrete surfaces sheltered from severe rain or freezing whilst wet Concrete exposed to condensation and rain Concrete continuously under water Concrete in contact or buried under non-aggressive soil/ground water Concrete surfaces sheltered from saturated salt air in coastal area. Concrete surfaces exposed to severe rain, alternate wetting and drying or occasional freezing whilst wet or severe condensation. Concrete completely immersed in sea water Concrete exposed to coastal environment. Though chlorine (or Bleaching Powder) is used for chlorinating water in water storage tanks, its quantity is controlled to maintain the quality of water to remain potable and fit for human consumption. In view of the above, the exposure condition proposed as severe is not justified. It should be changed to moderate. Excessive cover may enhance corrosion due to broader crack-widths at the tensile face of concrete surfaces.
2. IS: 3370 (Part 2)
Clause 4.4 Limit State Design is totally vague, as relevant details about LSD are not given. Its sub-clause 4.4.1.1refers to clause 35.2 of IS 456: 2000, which does not cover this type of structures. Clause 4.4.2 (c) makes no sense in case of Limit State Design. Clause 4.4.3.1 for limiting the crack width should be modified. It is important to provide smaller dia bars at closer spacing. A limit of 150 mm on the maximum spacing of hoops or meridional bars on tension face in water container portion may be included.

Severe

Comments

13

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Permissible stress in direct tension in reinforcing bars be limited to 0.36fy. Permissible stress in bending tension on water retaining face may be limited to 0.36fy and to 0.46fy on the other face. Table 4 should be suitably modified. Permissible stresses for M 25 concrete may be included in Table 1, 2 & 3. Recommendations for limiting the maximum water head in containers of overhead tanks should be included. The limit on maximum water-head over the bottom ring beam may be kept to 6.0 m for capacities up to 1000 kL and to 8.0 m for capacities above 1000 kL. Recommendations for minimum thickness to be adopted may be included. The proposed values for thickness of various components are as below:

Top dome or roof slab 100 mm Side walls 120 mm Bottom dome or bottom slab 120 mm Conical dome in Intze tanks 120 mm
Clause 8.1 (containing recommendations about minimum reinforcement) has typing mistakes. Moreover, for circular or Intze tanks having the tank diameter greater than 15.0 m, percentage of reinforcement in components of container appears to be increased too drastically. For example according to existing code, for a shell having 450 mm thickness, minimum amount of reinforcement required for HSD bars is 0.16 %, while according to this clause, it will be 0.36 %. Since, no undesirable effect due to the old provisions has been reported, this clause should be suitably altered to bring back the minimum percentage of reinforcement closer to old values. Clauses for thickness of shaft in shaft-type staging, buckling criterion for shafts, range giving the ratio of tank dia and shaft dia, minimum percentage of reinforcement, maximum spacing of reinforcement on each face of shaft and detailing around openings should be included.

3. IS: 1893 (Part 2)

For Seismic zone III IV

Lateral shear due to seismic force (kN) As per IS: 1893 - 1984 193.7 242.1 Proposed draft of IS 1893 386.5 579.8

As discussed earlier, none of the tank has failed due to earthquake. Why such a drastic change is proposed? In view of the above, the following parameters needs re-evaluation Two mass model for estimating seismic force Values of R, I & Z

Sushil K. Agarwal Former Professor, IIT Roorkee 14

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Appendix B Excerpts from the paper by Dr. Vipul Prakash Importance Factor

In authors opinion, the value of Importance Factor for tanks used for storing drinking water should be 1.0 rather than 1.5, because of the following reasons: 1. The scope for direct loss of life due to collapse of overhead tanks is practically nil. This is in contrast to the significant scope for direct loss of life due to collapse of ordinary buildings. 2. In India, the scope for indirect loss of life due to collapse of overhead tanks is also nil. Fires do not occur after earthquakes in India, because cooking gas for domestic use is supplied in cylinders rather than through pipelines and houses/buildings are constructed in masonry or concrete rather than in flammable timber. Every household in India stores water for drinking. In case of scarcity water is routinely supplied by tankers in most cities and towns. Following an earthquake, if the water tanks are inoperable for any reason, then water can be supplied through pumping and motor-driven tankers. If water is unavailable, then the overhead water tanks are likely to be empty and consequently unlikely to get even cosmetically damaged during an earthquake. 3. Thus failure of a water tank during an earthquake poses no risk to life and has negligible consequences for the environment, society and economy. Therefore, there is no reason to specify an Importance Factor, I = 1.5, for water tanks; and, I = 1.0 is appropriate.
Response Reduction Factor R

The following justification appears in the proposed draft IS:1893 (Part 2):2002 for assigning much lower values of R for LRS compared to buildings,
Response reduction factor (R), represents ratio of maximum seismic force on a structure during specified ground motion if it were to remain elastic to the design seismic force. Thus, actual seismic forces are reduced by a factor R to obtain design forces. This reduction depends on over strength, redundancy, and ductility of structure. Generally, liquid containing tanks posses low over strength, redundancy, and ductility as compared to buildings. In buildings, non structural components substantially contribute to over strength; in tanks, such non structural components are not present. Buildings with frame type structures have high redundancy; ground supported tanks and elevated tanks with shaft type staging have comparatively low redundancy. Moreover, due to presence of non structural elements like masonry walls, energy absorbing capacity of buildings is much higher than that of tanks. Based on these considerations, value of R for tanks needs to be lower than that for buildings. All the international codes

15

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

specify much lower values of R for tanks than those for buildings. As an example, values of R used in IBC 2000 are shown in Table C-2. It is seen that for a building with special moment resisting frame value of R is 8.0 whereas, for an elevated tank on frame type staging (i.e., braced legs), value of R is 3.0. Further, it may also be noted that value of R for tanks varies from 3.0 to 1.5. Values of R given in the present guideline (Table 2) are based on studies of Jaiswal et al. (2004a, 2004b). In this study, an exhaustive review of response reduction factors used in various international codes is presented. In Table 2, the highest value of R is 2.5 and lowest value is 1.3. The rationale behind these values of R can be seen from Figures C-4a and C-4b. In Figure C-4a, base shear coefficients (i.e., ratio of lateral seismic force to weight) obtained from IBC 2000 and IS 1893 (Part 1) 2002 is compared for a building with special moment resisting frame. This comparison is done for the most severe seismic zone of IBC 2000 and IS 1893 (Part 1: 2002. It is seen that base shear coefficient from IS 1893 (Part 1): 2002 and IBC 2000 compare well, particularly up to time period of 1.7 sec. In Figure C-4b, base shear coefficient for tanks is compared. This comparison is done for the highest as well as lowest value of R from IBC 2000 and present code. It is seen that base shear coefficient match well for highest and lowest value of R. Thus, the specified values of R are quite reasonable and in line with international practices. Elevated tanks are inverted pendulum type structures and hence, moment resisting frames being used in staging of these tanks are assigned much smaller R values than moment resisting frames of building and industrial frames. For elevated tanks on frame type staging, response reduction factor is R = 2.5 and for elevated tanks on RC shaft, R = 1.8. Lower value of R for RC shaft is due to its low redundancy and poor ductility (Zahn, 1999; Rai 2002).

In authors opinion each of the above justifications for low values of R for tanks compared to buildings can be questioned. Let us examine each justification one by one: 1. Response reduction factor (R) represents ratio of maximum seismic force on a structure during specified ground motion if it were to remain elastic to the design seismic force. Thus, actual seismic forces are reduced by a factor R to obtain design forces. This reduction depends on over strength, redundancy, and ductility of structure. The factor 2R should be thought of merely as a calibration factor to obtain the desired seismic force for design. Historically low-rise ordinary structures that are simple, symmetric and regular in plan, regular in elevation, and designed for a seismic coefficient of 0.08 to 0.12 (depending upon the site-soil conditions as given in IS 1893:1962) have performed well (i.e., with acceptable damage) when subjected to strong ground motions. Therefore, no matter what the definition of MCE, the final value of seismic coefficient for design of ordinary buildings must be brought to an acceptable value between 0.08 and 0.12 in the severest seismic zone. To get a seismic coefficient of 0.08 for hard soils, 0.096 for average soils, 16

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

and 0.12 for soft soils, this factor with a value of 12.5 was implicitly built-in the S curve for a in IS 1893:1984. In IS 1893 (Part 1):2002, this factor has a value of g 10 so as to obtain a seismic coefficient value of 0.09 for all soil types. Owing to increased strong motion instrumentation, values of Peak Ground Accelerations (PGAs) well in excess of 0.08g have been recorded during the past 65 years. Even before this instrumentation, during the 1897 Assam Earthquake witnesses cited phenomenon that would require PGAs greater than 1.0g. PGAs in excess of 0.36g continue to be recorded in the world. Therefore, Z = 0.36 , implying a PGA of 0.36g for MCE in the severest seismic zone, Zone V, of India is arbitrary. In IS 1893 (Part 1):2002, MCE is arbitrarily defined as, The most severe earthquake effects considered by this standard. Higher values of Z could as well have been considered for adoption, because PGAs greater than 1.0g have been recorded, but then it would have been difficult to offer a rational explanation for the required Response Reduction Factor. Even, with Z = 0.36 , it was necessary to split the reduction factor into two parts, a factor 2 to reduce MCE to DBE (Design Basis Earthquake) level and a factor R to further reduce it for use in design. Just like MCE, DBE is also arbitrarily defined and means nothing, because the seismic coefficient for DBE is not really used for design. For use in design, it is further divided by a factor R, for which a maximum value of around 5 is justifiable on the basis of maximum displacement ductility achievable in a structure. The value of R = 3 for OMRF buildings was then specified on the basis of calibration with IS 1893:1984 so as to obtain about the same values of design seismic coefficient for OMRF buildings. Z I Sa was merely 2 R g a calibration factor based on actual performance of a class of structures during earthquakes, then there would have been no difficulty in bringing out parts 2, 3, 4 and 5 of IS 1893 in 2002 itself. In fact there would not have been any need for splitting IS 1893 fifth revision into five parts. Problems arose, because 2R was not recognized as merely being a calibration factor, and rational explanations were sought for this factor. Therefore, to postpone the problem of identifying rational explanations for this factor for structure other than buildings, it was decided to split IS 1893 and agree to only Part 1 in 2002. If it were recognized and accepted that the factor 2R in Ah = For the case of buildings, the ill-defined terms like over-strength, redundancy and ductility provided the rationalization. In authors opinion, the over-strength is already accounted for by the term 2 in the denominator, and should not be counted again as part of the factor R. If this opinion is incorrect, then in what way is MCE being considered? If redundancy is taken to mean degree of static indeterminacy, then a conscientious designer would have taken its benefits in reducing the size of structural members, and it is therefore illogical to consider redundancy as a part 17

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

of the factor R . If redundancy is taken to mean number of members resisting a load, then as per one interpretation a cylindrical shell staging comprises only one member and is therefore non-redundant and deserves a low value of R ; whereas, as per another interpretation a shell staging comprises infinite members, and therefore, infinitely redundant and deserves a high value of R . Therefore, redundancy is clearly an ill-defined term. Thus only one term, ductility, remains to account for the value of R , and the value of R = 3 for OMRF buildings cannot be justified on the basis of ductility, because OMRF buildings dont have much ductility, and would require R 1 . It is interesting to know that ATC 19 (1995) committee notes in its first concluding remark that, There is no mathematical basis for the response modification factors tabulated in modern seismic codes in the United States. 2. Generally, liquid containing tanks posses low over strength, redundancy, and ductility as compared to buildings. In buildings, non structural components substantially contribute to over strength; in tanks, such non structural components are not present. Buildings with frame type structures have high redundancy; ground supported tanks and elevated tanks with shaft type staging have comparatively low redundancy. Moreover, due to presence of non structural elements like masonry walls, energy absorbing capacity of buildings is much higher than that of tanks. Based on these considerations, value of R for tanks needs to be lower than that for buildings. As explained in point 1 above, the over-strength is already accounted for by the factor 2 and redundancy is ill-defined, and its benefit already taken by the conscientious designer. It is, however, true that masonry walls can contribute to over-strength and energy-dissipation. But, the presence of masonry walls is not required by the code, and indeed they may not be present where they are most required, for example in the ground storey in a multi-storey construction. Masonry walls can also be harmful increase the eccentricity and introduce torsion in buildings. Therefore, most earthquake engineers recommend that masonry walls be either effectively isolated from the structural system or be effectively integrated with the structural system. When isolated, the masonry walls can neither provide the possible benefits nor the possible drawbacks. And when integrated, the masonry walls are considered as structural members and their benefits and drawbacks are properly accounted for, and therefore, it is illogical to count masonry walls as contributing to high value of R for buildings. On the other hand, value of R for ground reservoirs and overhead tanks could be much higher than that for buildings, because the tanks and their staging are simple, symmetric in plan, regular in plan and elevation, designed by competent engineers, constructed by experienced contractors, subjected to more intense inspections during construction, unlikely to ever be modified or undergo change of use, and subjected to their maximum normal design loads on a daily basis. All these factors contribute to their high construction quality and dependability compared to buildings.

18

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

3. All the international codes specify much lower values of R for tanks than those for buildings. Thus, the specified values of R are quite reasonable and in line with international practices. By international practices the practices in USA are mainly being referred to. The US code, IBC 2000, being referred to is called International Building Code 2000, even though there is nothing international about it. In USA, different states have traditionally evolved and followed different codes. IBC is to be followed by all states of US and hence the word, International in the title. However, let us examine the recommendation of IBC 2000. The R values in IBC 2000 reported by Jain et al. (2004) are contained in Table C-2 reproduced below.

19

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

In the above table, kindly note the following: (a) R = 8 for buildings with SMRF and R = 3 for buildings with OMRF. This is in contrast to the values R = 5 for SMRF and R = 3 in IS 1893 (Part 1):2002. (b) For elevated tanks staging with SMRF are not specified. The reason for this is that for SMRF the design seismic loads are lower and reliance on ductility is higher. Higher ductility implies greater lateral drifts and greater damage, which is not acceptable for elevated water tanks, which are considered as essential structures in US, because of the fire hazard there. IS 1893 (Part 1):2002 and IS 13920:1993, however, take the opposite view. They prohibit the construction of buildings with OMRF and Ordinary Shear Wall Frames in seismic Zones III, IV and V; and permit only buildings with SMRF and Ductile Shear Wall Frames. In authors opinion it is wrong to prohibit any structural system in any seismic zone. What is the logic of prohibiting an OMRF building that is designed for elastic response under MCE or for DBE? Similarly, what is the logic of prohibiting a single storied OMRF building with masonry walls, when a building with masonry walls alone is permitted? Instead the code should specify higher seismic forces for design of vulnerable building types. In fact, buildings designed for high ductility would have lower lateral stiffness and strengths and hence should sustain greater damage. Conversely, buildings designed for low ductility would have higher lateral strength and stiffness and hence should sustain less damage. Therefore, IBC 2000 does the right thing in allowing both OMRF and SMRF for staging of elevated tanks, but specifying the same R=3 for both. In practice, economy will dictate that the staging be of OMRF type. (c) Among elevated tanks the following types are listed:

Note that R = 3 is used for elevated tanks supported on braced/unbraced legs as well as for those supported on structural towers similar to buildings. Note that R = 3 is also specified for buildings with OMRF. Note also that elevated tanks supported on single pedestal for which R = 2 is specified, are unlikely to be similar to elevated tanks on cylindrical shaft staging. In US practice, the single pedestal is likely to be very narrow,

20

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

whereas in India the cylindrical shaft staging is as wide as the comparable columnar staging. Now examine the type of staging and R values specified in Table 2 of proposed draft IS 1893 (Part 2), reproduced below:

Clearly if a fair comparison was made the R value for Tank supported on OMRF RC frame should have been equal to 3.0, rather than 1.8. If the R values in Table 2 above are increased in the ratio 3.0/1.8, then the new values will be as follows:
Type of Tank Tank supported on masonry shaft a) Masonry shaft reinforced with horizontal bands b) Masonry shaft reinforced with horizontal bands and vertical bars at corners and jambs of openings Tank supported on RC shaft RC shaft with two curtains of reinforcement, each having horizontal and vertical reinforcement Tank supported on RC frame a) Frame not conforming to ductile detailing, i.e., ordinary moment resisting frame (OMRF) b) Frame conforming to ductile detailing, i.e., special moment resisting frame (SMRF) Tank supported on steel frame R 2.17 2.5 3.0

3.0 4.17 4.17

In fact, R values more consistent with values given in IS 1893 (Part 1):2002 for buildings could be chosen as follows.

21

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Elevated Tank Type of Tank Tank supported on masonry shaft a) Masonry shaft reinforced with horizontal bands b) Masonry shaft reinforced with horizontal bands and vertical bars at corners and jambs of openings Tank supported on RC shaft RC shaft with two curtains of reinforcement, each having horizontal and vertical reinforcement Tank supported on RC frame a) Frame not conforming to ductile detailing, i.e., ordinary moment resisting frame (OMRF) b) Frame conforming to ductile detailing, i.e., special moment resisting frame (SMRF) Tank supported on steel frame R 2.5 3.0 3.0

3.0 5.0 5.0

Likewise the R values proposed for ground supported tanks in draft IS 1893 (Part 2) are too low and a uniform value R = 3 should be adopted regardless of the base and walls. The quality of construction in tanks even masonry tanks is bound to be far better than the quality of construction of walls in buildings, because a tank is subjected to the full normal loads on a daily basis. 4. Elevated tanks are inverted pendulum type structures and hence, moment resisting frames being used in staging of these tanks are assigned much smaller R values than moment resisting frames of building and industrial frames. For elevated tanks on frame type staging, response reduction factor is R = 2.5 and for elevated tanks on RC shaft, R = 1.8. Lower value of R for RC shaft is due to its low redundancy and poor ductility (Zahn, 1999; Rai 2002). Elevated tanks are not inverted pendulum type structures. An inverted pendulum is an unstable structure, whereas elevated tanks are very stable structures. A pendulum hangs from a hinge, whereas an elevated tank is fixed at its both ends and at bracing levels. Its behaviour is better than a multi-storied building, and more akin to that of a single storied building. In an elevated tank, the design axial, shear and bending moment in all members of the staging (columns and braces) remains about the same, just like in a single-storied building. In a multi-story building the story shear and column axial forces reduce with height. If the graphical description, inverted pendulum is accepted for an elevated tank, then it should be equally acceptable for a single story building, which also has a single heavy mass situated at the top. With the suggested changes in the Importance Factor and Response Reduction Factor, the values of the design horizontal seismic coefficient for tanks shall change only marginally from the values in IS 1893:1984. Only a marginal change in values is warranted as the existing water tanks have exhibited exemplary performance in earthquakes in India.

22

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Appendix C

List of Delegates
Design Experts: S. No. 1. Name & Designation Dr. Anand Prakash,

Ex-Professor, Department of Civil Engineering, I.I.T. Roorkee, Roorkee. 2. Prof. S.K. Agarwal, Ex-Professor, Department of Civil Engineering, I.I.T. Roorkee, Roorkee.
Delegate Field Engineers:

Uttar Pradesh
S. No. 3. Name & Designation Er. A. K. Srivastava, Chief Engineer, (Rural), U.P. Jal Nigam, Lucknow. Er. Vijendra Vikramaditya, Superintending Engineer, 18th Circle, U.P. Jal Nigam, Ghaziabad. Er. Ashok Kumar, Deputy Materials Manager, U.P. Jal Nigam, Lucknow Er. R. K. Dwivedi, Executive Engineer, Const. Division, U.P. Jal Nigam, Jaunpur. Er. L. K. Gupta, Executive Engineer, Const. Division, U.P. Jal Nigam, Mathura. Er. Mohd. Aslam, Assistant Engineer, (Jhansi zone), U.P. Jal Nigam, Jhansi.

4. 5. 6. 7. 8.

Haryana 9. Er. D. R. Yadav Suprintending Engineer PWD Public Health Circle (Haryana) Karnal

23

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Punjab 10. Er. B.K Garg, Executive Engineer, Punjab Water and Sewerage Board Chandigarh

Uttaranchal 11. 12. 13. 14. 15. Er. R. N Verma, Superintending Engineer, IXth Circle, Uttaranchal Peyjal Nigam, Dehradun. Er. S. K. Semwal, Manager Appraisal, Head Office, Uttaranchal Peyjal Nigam, Dehradun. Er. S. K. Agarwal, Executive Engineer, Doon Division, Uttaranchal Peyjal Nigam, Dehradun. Er. Avdhesh Kumar, Executive Engineer, Head Office, Uttaranchal Peyjal Nigam, Dehradun. Er. B. K. Jain, Executive Engineer, Central Store Division, Uttaranchal Peyjal Nigam, Dehradun. Er. M. K. Gupta, Executive Engineer, Head Office, Uttaranchal Peyjal Nigam, Dehradun.

16.

Structural Engineering Faculty at IIT Roorkee:

16. 17. 18. 19. 20.

Dr. Ashok K. Jain Prof. Of Civil Engineering I.I.T. Roorkee, Roorkee Dr. K.K. Singh Prof. Of Civil Engineering I.I.T. Roorkee, Roorkee Dr. Pradeep Bahrgava Associate Prof. Of Civil Engineering I.I.T. Roorkee, Roorkee Dr. Vipul Prakash Assistant Prof. Of Civil Engineering I.I.T. Roorkee, Roorkee Dr. Bhupinder Singh Assistant Prof. Of Civil Engineering I.I.T. Roorkee, Roorkee 24

Recommendations of the Workshop on Revision of IS Codes on LRS-IIT Roorkee

Registered Student Delegates from IIT Roorkee: S.No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 Name Jikum Hiri B. Sudharshan Reddy Durga Suneel Chalapaka Ramanjaneyulu B. Sadhu Venkata Rajesh K. Girish Babu Amol Arvind Mankar Pisal Yogesh Dattatraya S. Vamsidhar B. V. Lokesh Ajit Singh Venkata Kishor S. M.V.S. Ravikumar Pranay Vasantrao Urewar Debjyoti Das Ratheesh Kumar M.V. Kumar Satyam Anil Mishra Jyoti Prasad Jagtap Amit Cahndra Arijit Bhakat Ganesh L. Konar Surender Kumar Verma Bharmal Husen Ismaeel Sadaquat Ali Tarun Dandotiya Sachin M. Pore Krantikumar Boragaonkar Mohd Shariq Mohd Shahiq Khan Dipankar Das P.V.Mayur Babu Tesfaye Alemu V.Giri E.V.P.Bhanu Prakash R.Dileep Kumar Reddy Dinesh Kumar Jain Ajit Kumar P. Jayachandra Specialization M.Tech. 2nd Year Structural Engineering M.Tech. 2nd Year Building Science and Technology M.Tech. 2nd Year Structural Dynamics M.Tech. 2nd Year Structural Engineering M.Tech. 2nd Year Building Science and Technology M.Tech. 1st Year Building Science and Technology M.Tech. 2nd Year Structural Engineering M.Tech. 2nd Year Structural Dynamics M.Tech. 2nd Year Structural Engineering M.Tech. 2nd Year Geotechnical Engineering M.Tech. 2nd Year Structural Engineering M.Tech. 2nd Year Structural Engineering M.Tech. 2nd Year Structural Engineering M.Tech. 2nd Year Structural Engineering M.Tech. 1st Year Structural Engineering M.Tech. 1st Year Structural Engineering M.Tech. 1st Year Structural Engineering M.Tech. 1st Year Structural Engineering M.Tech. 1st Year Structural Engineering M.Tech. 1st Year Structural Engineering M.Tech. 2nd Year Geotechnical Engineering M.Tech. 1st Year Structural Engineering M.Tech. 1st Year Structural Engineering M.Tech. 2nd Year Building Science and Technology M.Tech. 2nd Year Building Science and Technology M.Tech. 2nd Year Structural Engineering Ph.D. Scholar Structural Dynamics Ph.D. Scholar Structural Dynamics Ph.D. Scholar Structural Engineering Ph.D. Scholar Structural Engineering M.Tech. 2nd Year Structural Dynamics M.Tech. 2nd Year Structural Engineering M.Tech. 2nd Year Computer Aided Design M.Tech. 2nd Year Computer Aided Design M.Tech. 2nd Year Environmental Engineering M.Tech. 2nd Year Transportation Engineering Ph.D. Scholar Structural Engineering Ph.D. Scholar Structural Engineering M.Tech 2nd Year-Building Science & Tech.

25