TG 10j SOUTH AUSTRALIAN WATER CORPORATION

Transcription

1 TG 10j SOUTH AUSTRALIAN WATER CORPORATION TECHNICAL GUIDELINE GENERAL TECHNICAL INFORMATION FOR GEOTECHNICAL DESIGN ~ Part J ~ Pipelines Issued by: Manager Engineering Issue Date:

2 SA Water 2007 This document is copyright and all rights are reserved by SA Water. No part may be reproduced, copied or transmitted in any form or by any means without the express written permission of SA Water. The information contained in these Guidelines is strictly for the private use of the intended recipient in relation to works or projects of SA Water. These Guidelines have been prepared for SA Water s own internal use and SA Water makes no representation as to the quality, accuracy or suitability of the information for any other purpose. It is the responsibility of the users of these Guidelines to ensure that the application of information is appropriate and that any designs based on these Guidelines are fit for SA Water s purposes and comply with all relevant Australian Standards, Acts and regulations. Users of these Guidelines accept sole responsibility for interpretation and use of the information contained in these Guidelines. SA Water and its officers accept no liability for any loss or damage caused by reliance on these Guidelines whether caused by error, omission, misdirection, misstatement, misinterpretation or negligence of SA Water. Users should independently verify the accuracy, fitness for purpose and application of information contained in these Guidelines. The currency of these Guidelines should be checked prior to use. No Changes Required In the January 2007 Edition The following lists the major changes to the December 2004 edition of TG 10j: 1. Nil 2 of 52

3 Contents SA WATER NO CHANGES REQUIRED IN THE JANUARY 2007 EDITION... 2 TABLES & FIGURES... 4 SECTION 1: SCOPE... 6 SECTION 2: PIPE ANCHOR BLOCK OUTLETS... 6 SECTION 3: TEMPORARY FIRE PLUG END ANCHORS Geotechnical Recommendations Further Geotechnical Recommendations Dimensions for In-Line Stop Valves & Temporary Dead End Fireplugs Anchor Block Dimensions for In-Line Stop Valves & Temporary Dead End Fireplugs 11 SECTION 4: VALVE ANCHOR DESIGN Allowable Horizontal Bearing Pressure for The Ground Anchor Design (See Following Diagram)...12 SECTION 5: EMBEDMENT HOW IT WORKS What Native Soil Modulus to Design For What Embedment Soil Modulus to Design For...17 SECTION 6: EMBEDMENT SAND SIMPLE FIELD TEST Test Procedure...18 SECTION 7: THE METHOD SPECIFICATION FOR EMBEDMENT COMPACTION Discussion...22 SECTION 8: EMBEDMENT OVERLAY THICKNESS FOR SEWER...23 SECTION 9: EMBEDMENT COMPACTION ELASTIC MODULUS VS DENSITY...24 SECTION 10: CONTROLLED LOW-STRENGTH MATERIAL (CSLM) FOR EMBEDMENTS The Leeds clsm Embedment Trial...27 SECTION 11: DS4 REVIEW...30 SECTION 12: PIPE LAYING GROUNDWATER CONTROL SPECIFICATION CLAUSES...32 SECTION 13: PIPE TRENCH WIDTH DISCUSSION...32 SECTION 14: PIPE LAYING GROUNDWATER CONTROL USING SCREENINGS...33 SECTION 15: PIPE LAYING IN EMBANKMENTS...34 SECTION 16: PIPE TRENCH EXCAVATION STABILITY SAMPLE LOG SHEET...36 SECTION 17: PIPE LAYING ROADS OVER OLD MAINS...37 SECTION 18: PIPE LAYING IN ROADS COMPACTION REQUIREMENTS Trench Floor Preparation Bedding Placement Side Support & Overlay Placement & Compaction of 52

4 18.4 Table of Minimum Densities (%)...38 SECTION 19: PIPELINE HYDRAULIC TEST SPECIFICATION General Filling the Pipeline Pre-Conditioning Concrete Lined Pipes Test Pressure Pressure Test Reports Emptying the Pipeline Hydraulic Testing of Pipework (For HOCV Installations Etc)...42 SECTION 20: SEWER SCREENINGS 10-7MM VS 75MM...43 SECTION 21: SEWER SCREENINGS REVIEW OF SPECIFICATION Background on the Use of Screenings for Sewer Embedment The Proposal to Add 14 Mm Aggregate to the Specification Recommendation...46 SECTION 22: LAYING SEWERS IN REACTIVE SOILS...47 SECTION 23: SEWER MAINTENANCE SHAFT RAISERS Analytical Model Formula for the Normal Force on The River Murray Water (MDBC) Formula for the Down-Drag Force on The Riser Calculation of Typical Down-Drag Forces Other Considerations Conclusions & Recommendations...51 SECTION 24: SEWER MAINTENANCE SHAFT RAISERS Tables & Figures Table 3.1 Recommended Distance of Fireplug Thrustblock from Open Channel... 8 Table 3.2 Minimum Length of Trench Downstream of Fireplug Dead End Anchor... 9 Table 3.3 Dimensions for In-Line Stop Valves & Temporary Dead End Fireplugs Table 3.4 Anchor Block dimensions for In-line Stop Valves & Temporary Dead-end Fireplugs Table 4.1 Thrust Collar standard details Table 6.1 Embedment Sand Sample Test Results Table 9.1 Field Identification Tests Table 18.1 Table of Minimum Densities (%) Table 21.1 Coarse Aggregate Grading Requirements of 52

5 Figure 4.1 Illustration of Anchor Design Figure 5.1 Pipe Deflection Under Loads Figure 5.2 Reduced Deflection of Pipe Embedded in Soil Figure 5.3 The Spring Analogy for Side Support Figure 5.4 Pipe Embedment Stiffness (or Modulus) Figure 5.5 Illustration of Why Side Support Must be Uniform Figure 5.6 Minimum Practical Limit for Density of Pipe Embedment Figure 6.1 A Simple Field Acceptance Test for Pipe Embedment Sand Figure 10.1 The CLSM Embedment Trial Set-Up Figure 23.1 Illustration of the Potential Down-drag Force on the Riser Figure 23.2 Diagram of Forces on the Riser of 52

6 Section 1: Scope Section 2: Pipe Anchor Block Outlets Sometimes it is considered necessary to find alternative anchoring systems for block outlets, in particular whether the valve could be braced to the wall of the sump. The sump is unlikely to be able to resist the thrust, particularly with the bigger pipes (the drawings you gave showed that up to 200 diameter may be used). Such sumps are generally designed to resist only light, uniform, external loads which put the sump wall into compression. They are not designed to resist local internal thrust. The situation is made worse by the fact that the thrust would be near the base of the sump and that the base is unrestrained. There would also be the difficulty of ensuring adequate compaction of the backfill outside of the sump. Two alternatives appear possible: 1. Make the MSCL special longer so that the anchor can be located well away from the disturbed ground near both the sump and the growers connection. 2. The main disadvantage would be that even with this approach it might still be difficult to find undisturbed or sufficiently strong/dense ground. 3. Use a cast-in-place base slab for the sump, and design the base slab to act as the anchor. The valve would be bolted to the slab on a standard pedestal. The sump would simply sit on the base slab. This approach would have the advantage that the anchor would simply be a mass of unreinforced concrete poured into fresh excavation prior to any pipelaying. This would provide the best possible conditions for anchoring (deep, confined and undisturbed), simplify the MSCL special, require no formwork, and require no special care to be taken during the pipe trench excavation and backfilling. This Technical Note was prepared by Ed Collingham, 28/05/1999 (Ex Principal Engineer Geotechnical) Section 3: Temporary Fire Plug End Anchors 3.1 Geotechnical Recommendations Below are the recommendations put forward by Ed Collingham, Ex Principal Engineer concerning temporary fireplug dead ends and anchors. 6 of 52

7 THE SITUATION A temporary fireplug dead end requires a thrust block. The thrust block usually bears on a puddle flange around a specially cast extension pipe immediately upstream of the fireplug. When the trench downstream of the fireplug is reopened (to continue the main) the main is usually pressurised. If the distance between the thrust block and the open trench is too short there is a risk that the soil stresses from the block will cause the walls of the open trench to collapse inwards. SOLUTIONS It would be possible to come up with a set of special thrust block designs to eliminate this risk. They would need to take into account pipe size, design head, trench width, trench wall soil type etc. They would be considerably larger than standard thrust blocks for dead ends. The recommended alternative solution is to ensure that the thrust block is sufficiently far upstream of the open trench that the stresses in the soil have effectively dissipated. A standard thrust block can then be used. RECOMMENDATION I have calculated this distance required between the thrust block and the open trench for 100, 150 and 200 mm diameter mains based on: A distance between the thrust block and the open trench of three times the width of the bearing area of the thrust block on undisturbed natural soil. Thrust block dimensions appropriate for a design head of 140 m, a trench wall soil safe bearing capacity of 50 kpa, and the shape restrictions to be recommended for standard thrust blocks in the Blue Book. The assumption that the fireplug will also be uncovered while the connection is being made. The assumption that the upstream face of the anchor block is as close as possible to the upstream flange of the special (ie about 150 mm). With these criteria, and for these pipe sizes, the required safe distance can be achieved by casting the anchor around one maximum standard length extension pipe, and adding one more maximum standard length extension between it and the fireplug. Details are tabulated below: 7 of 52

8 Table 3.1: Recommended Distance of Fireplug Thrustblock from Open Channel. Main Diameter Required Distance (mm) Recommended (mm) Extension Pipes for anchor plus 1 x for anchor plus 1 x for anchor plus 1 x 1065 Available Distance (mm) (near enough) 3.2 Further Geotechnical Recommendations Below are further recommendations put forward by Ed Collingham, Ex Principal Engineer concerning temporary fireplug dead ends and anchors. BACKGROUND The Water Supply Construction Manual and DS9 are currently being upgraded by the Water Systems section of ESG. I am providing the geotechnical input to this upgrade. Your formal approval is requested for a change in the installation procedure for anchors at all in-line stop valves and temporary fireplug dead ends. All stop valves and temporary fireplug dead ends on rubber-ring joint pipes require an anchor. The anchor is usually poured around a standard-length extension pipe specially cast with a puddle flange. Bolted flanged joints are therefore found immediately upstream and downstream of the anchor. The current DS9 calls for all joints on a main to be left exposed during the pressure test to allow them to be inspected for leakage. This means that the trench must be left open downstream of the anchor to allow inspection of the joint. Geotechnically, it is not possible to design an anchor of acceptable size or shape (except for the smallest main diameters) to cope with an unsupported open trench downstream of it. The Principal Engineer Pipelines and Structures, D G Kerry, in his minute dated , makes recommendations for pressure testing which require that all of the embedment (including that over the joints) be placed before the test (ie to 150 mm above the top of the pipe). While I concur with his recommendations, the placement of the embedment alone would not significantly improve the support given to an anchor. The trench downstream of any anchor must therefore be either strutted or completely backfilled to the surface with properly compacted material during the pressure test and under operating conditions. From our discussions on I understand that you prefer complete backfilling instead of strutting on the grounds that strutting introduces a new technique which may not be done adequately or at all. I concur. For in-line stop valves (which would not normally be exhumed without first being isolated) it will be sufficient to use the standard puddle flanged extension pipe and to specify that the trench be fully backfilled for a certain distance downstream of the anchor prior to testing. 8 of 52

9 For temporary fireplug dead ends however, it is expected that the trench downstream will be reopened while the main is in service (ie when the time comes to connect in the extension main). We therefore need to ensure that the anchor is sufficiently far upstream of the open trench that the stresses in the soil have effectively dissipated. A standard anchor can then be used. RECOMMENDATIONS 1. That the required minimum length of trench to be kept backfilled downstream of a temporary fireplug dead end anchor be achieved by using a long fabricated mild steel special, which could include both the puddle flange and the fireplug tee, and thus reduce rather than increase the number of bolted joints (as discussed). 2. That the minimum lengths be as follows: Table 3.2: Minimum Length of Trench Downstream of Fireplug Dead End Anchor. Nominal Pipe Diameter Minimum Length of Trench to be Kept Backfilled Downstream of a Temporary Fireplug Dead End Anchor 100 mm 1200 mm 150 mm 1600 mm 200 mm 2100 mm These lengths are based on: Anchor dimensions appropriate for a test head of 140 m (in accordance with DGK recommendations), a trench wall soil allowable horizontal bearing pressure (AHBP) of 50 kpa, and the shape restrictions to be recommended for thrust blocks and anchors in the Water Supply Manual. A distance between the downstream face of the anchor and the open trench of between three and four times the width of the bearing area of the anchor on undisturbed natural soil (varies with shape of bearing area). 9 of 52

10 3.3 Dimensions for In-Line Stop Valves & Temporary Dead End Fireplugs Table 3.3: Dimensions for In-Line Stop Valves & Temporary Dead End Fireplugs. Nominal Pipe Diameter SOIL CLASSIFICATION (See Notes 1 and 2) Stiff Clay Medium Dense Clean Sand (50 kpa AHBP) (3) Very Stiff Clay Dense Clean Sand/Gravel Decomposed Rock (100 kpa AHBP) (3) Hard Clay Sound Rock (200 kpa AHBP) (3) ANCHOR BLOCK DIMENSIONS ANCHOR BLOCK DIMENSIONS ANCHOR BLOCK DIMENSIONS W mm H mm T mm W mm H mm T mm W mm H mm T mm 100 mm mm mm mm 2000 total mm Anchor exceeds max standard width: special design required or use Tyton-Lok joints * 375 mm * Reinforcing - Y16 bars both faces at 140 mm centres for all anchors, except 100 mm centres where indicated by *. 10 of 52

11 3.4 Anchor Block Dimensions for In-Line Stop Valves & Temporary Dead End Fireplugs Table 3.4: Anchor Block dimensions for In-line Stop Valves & Temporary Dead-end Fireplugs Nominal Pipe Diameter Very Soft, Soft or Firm Clay Loose Clean Sand Domestic Refuse Uncompacted Fill ANCHOR BLOCK DIMENSIONS W mm H mm T mm 100 mm Standard anchors cannot be used - arrange for geotechnical site investigation and special design or use Tyton-Lok joints. W mm H mm T mm 150 mm 200 mm 250 mm 300 mm 375 mm For domestic refuse and uncompacted fill, a geotechnical site investigation and special design is essential. NOTES: 1. For soil classification guidelines refer to drawing ###. 2. If the watertable is, or could rise, close to the surface then downgrade the soil classification to the next weakest soil on the left in the above table. 3. AHBP = Allowable Horizontal Bearing Pressure. 4. Total anchor block width = trench width plus 2 times W. 5. Maximum standard anchor block width = 2000 mm. 6. Trench must be fully backfilled or strutted to surface downstream of anchor block for a minimum distance of 4 times W during the pressure test. 7. Anchor blocks designed for a test pressure of 1.4 MPa (140 m head). 8. DICL pipe with Tyton-Lok joints may be used instead of anchor blocks. This Technical Note was prepared by Ed Collingham, 24/11/1993 (Ex Principal Engineer Geotechnical) 11 of 52

12 Section 4: Valve Anchor Design Below are the recommendations put forward by Ed Collingham, Ex Principal Engineer concerning the Woolpunda SIS-Bore 14 Valve Anchor Design. 4.1 Allowable Horizontal Bearing Pressure for The Ground Based on your photos and our discussion I would judge the ground at pipe depth to be equivalent to dense sand/gravel. For a normal Blue Book (BB) anchor design the allowable horizontal bearing pressure (AHBP) for such a soil would be 100 kpa. But note that the BB AHBPs are based on 750 mm pipe cover and keeping the pull-out at the first RRJ to less than 10 mm. At Bore 14 it is understood that the depth to the top of the pipe is 1400 mm, which is a depth increase to the centre line of the pipe of 70% over the BB standard. This would permit the AHBP to be increased somewhat. 4.2 Anchor Design (See Following Diagram) This anchor is a special design, not a standard BB design, because: 1. The head on the valve will not exceed 100 m (BB standard is 140 m). 2. The excavations for the pipe trench and the anchor have already been dug and are both over the normal BB size. (The anchor thickness is 600 mm compared to a BB standard of 350 mm, and the trench width is 1550 mm compared to a BB standard of 650 mm). 3. The thrust will only be in one direction. (The BB allows thrust from both directions.) Note that the extra thickness would give the anchor a great increase in its beam capacity more than compensating for the increased in span caused by the oversize trench width. The thrust on the valve would be about 90 kn (9 tonne). For an anchor height of 900 mm the total bearing area will be about 1.8 square metres, implying a horizontal bearing pressure of about 50 kpa which is comfortable. 12 of 52

13 Reinforcement Thrust Collar 600 Thrust Direction Reinforcement: Figure 1: Illustration of Anchor Design. Vertical reinforcement: Y12 bars at 150 c/c*. Horizontal reinforcement: Y16 bars at 140 c/c*. Ensure equal and maximum number of bars above and below pipes. Cut bars less than 300 long may be omitted. Minimum cover 75 mm. Use Grade 20 concrete. *Bar size and spacing can be adapted to materials on hand contact if alternatives preferred. Thrust Collar: As per Standard Drawing 75 2A (extract below): Table 4.1: Thrust Collar standard details. pipe plate thickness collar width collar thickness weld size* 5 mm 100 mm 12 mm 5 mm 6 mm 150 mm 16 mm 5 mm *The weld on the inside of the anchor must be a recessed fillet weld to present a flat thrust face to the concrete. The weld on the outside may be a normal fillet weld. This Technical Note was prepared by Ed Collingham, 21/08/2003 (Ex Principal Engineer Geotechnical) 13 of 52

14 Section 5: Embedment How it works VERTICALLY LOADED FREE PIPE allowable deflection typically 3% of D for pressure pipe D Figure 5.1: Pipe Deflection under Loads. A pipe loaded by trench fill and traffic will squash down vertically and bulge out sideways. The pipe itself could be designed to support the entire trench fill and traffic loads - but it would need to have a very thick wall. VERTICALLY LOADED PIPE WITH SIDE SUPPORT Figure 5.2: Reduced Deflection of Pipe Embedded in Soil. It is usually more efficient to rely on the material at the sides of the pipe to provide some lateral resistance to the bulging. This is known as side support. Side support reduces the vertical deflection for a given load and so allows a thinner pipe wall to be used. 14 of 52

15 Figure 5.3: The Spring Analogy for Side Support. The side support may be thought of as coming from elastic springs. The net side spring stiffness that the pipe feels depends on the stiffness of both the pipe embedment and the stiffness of the trench wall (the effective combined soil modulus ). Lower density Lower stiffness (or modulus) Higher density Higher stiffness or modulus) Figure 5.4: Pipe Embedment Stiffness (or Modulus). Usually little can be done about the stiffness of the trench wall material (the native soil modulus ), so the designer is limited to specifying the stiffness of the pipe embedment material (the embedment soil modulus ). The embedment soil modulus depends on the nature of the material (eg whether it is a silty sand or a gravel) and on its density. 15 of 52

16 Very low density Compresses with pipe Gives no side support High density Settles around pipe Side support retained Figure 5.5: Minimum Practical Limit for Density of Pipe Embedment. The density of the pipe embedment material must always be sufficient that it does not settle around the pipe under its self weight or under the trench fill or traffic loading. If it is loose enough to do this it is also too loose to provide proper side support. If side support is placed and compacted in one lift it will push the pipe down with it - defeating the purpose of the side support. Also it cannot be evenly dense. Place and compact side support in layers. Pipe stays round. Density is uniform. Figure 5.6: Illustration of Why Side Support Must be Uniform. 5.1 What Native Soil Modulus to Design For Native soil modulus is likely to be very variable. 16 of 52

17 Little can usually be done economically to improve it. It has less of an influence on pipe design than the embedment soil modulus. It is difficult to measure in the field. So it is generally reasonable to use a conservative value from established correlations with other soil parameters. Table 3.2 in AS/NZS :1998 gives correlations between soil type, density and modulus. Natural soils can have densities below the engineering range (ie less than say 90% of standard MDD), so Table 3.2 gives correlations for as low as 85% of standard MDD. 5.2 What Embedment Soil Modulus to Design For The embedment soil modulus can be controlled by specifying both the material and its density. Gives the designer flexibility to take account of the cost of materials, availability of materials, compaction methods, degree of site supervision, etc But it must be recognised that there is a minimum practical limit for embedment soil density, in that it must be at least sufficiently dense not to settle vertically down around the pipe under the loads from the trench fill or traffic. (See sheet 5) The embedment should also be sufficiently dense that it does not creep or permanently compress under the lateral stresses. (Note that soils appear to have quite a high elastic modulus under the very low-strain cyclic loading they are subjected to in laboratory modulus tests, but suffer creep under static loading, and permanent compression under slight overload.) Table 3.2 in AS/NZS :1998 (and section 3.4 generally) does not alert the inexperienced designer to these points and so could lead to inappropriate designs. There should be separate tables for native soils and compacted embedment materials in AS/NZS AS/NZS is a design standard. It should not be referenced in specifications for use by contractors. This Technical Note was prepared by Ed Collingham, 26/04/2000 (Ex Principal Engineer Geotechnical) 17 of 52

18 Section 6: Embedment Sand Simple Field Test Trench Fill Overlay Terminology used in pipelaying Embedment Fill Side Support Trench Floor Bedding 80 to 150 finished thickness Figure 6.1: A Simple Field Acceptance Test for Pipe Embedment Sand. The main requirements for a good pipe embedment sand are: 1. It will be easy to compact in the restricted area around a pipe. 2. Its ease of compaction will be relatively insensitive to moisture content. 3. If in contact with metal pipe or fittings it has a low specific conductivity. This simple field test described here checks only for the first two of these requirements ie whether the sand will be easy to compact and whether its ease of compaction will be sensitive to changes in moisture content. The test relies on the fact that a sand that is free-draining will normally also be easy to compact and that its compaction will not be sensitive to changes in moisture content. This test was found to correlate well with the normal grading based acceptance criterion for embedment sands namely that an embedment sand should be non-plastic and contain less than about 8% fines. ( Fines are defined as material less than 75 micrometre in diameter.) This test checks the sands on three different criteria drainage rate, liquefaction and penetration. The test procedure and a record sheet are presented below. 6.1 Test Procedure 18 of 52

19 Step 1 Preparation Take the sieve and check that the 75 micrometre mesh is firmly in place and undamaged. Step 2 Filling Fill loosely with the sand to be tested up to the lower lip. Step 3 Saturating Distribute 500 ml of water over the sand in one continuous, fairly quick, but smooth pour, taking care to keep the surface reasonably flat. Step 4 Drainage Rate Record (in seconds) how long it takes for the water on top of the sand to disappear. (Photographs of four examples are presented below.) 19 of 52

20 Step 5 Liquefaction Wait twenty seconds after the water has cleared from the surface, then gently tap the whole sieve on the ground five times. Record how many seconds it takes for any new water that appears to disappear again. (If none appears record zero.) Step 6 Penetration Wait at least another five minutes then press the ball of the thumb into the surface of the sand, using moderate pressure. Record whether or not the impression is more than 5 mm deep. The picture above-left shows an impression that is less than 5 mm deep. The surface felt firm. (This sand had 5% fines.) The picture left shows an impression that is greater than 5 mm deep. The full depth was completely sloppy. (This sand had 16% fines.) Drainage Example 1 This sand had 0% fines in it. It took less than 5 seconds for the water to clear the surface. 20 of 52

21 Drainage Example 2 This sand had 5% fines in it. It took about 45 seconds for the water to clear the surface. Drainage Example 3 This sand had 10% fines in it. It took about 90 seconds for the water to clear the surface. Drainage Example 4 This sand had 16% fines in it. Even after 3 hours the surface was still wet. Table 6.1: Embedment Sand Sample Test results. Sample Number Drainage Rate Time taken for initial water to disappear time in second s <90 s >90 s Liquefaction Time taken for new water to disappear (if none record zero) time in seconds <30 s >30 s Penetration Depth thumbprint depth in mm of <5 mm >5 mm Acceptance Three = OK Two or less = not OK OK >10 not OK 21 of 52

22 This Technical Note was prepared by Ed Collingham, 03/02/2002 (Ex Principal Engineer Geotechnical) Section 7: The Method Specification for Embedment Compaction Why use a method spec rather than a performance spec for the placement and compaction of pipe embedment sand? (A method spec says HOW a job should be done it spells out the steps that need to be followed to achieve the required engineering result. A performance spec says only WHAT the final result should look like but compliance is limited to things that can be measured.) It is important to use a method spec for the placement and compaction of pipe embedment sand because: 1. It is just as important HOW the required density is achieved in pipe embedment sand as it is that the required density is achieved at all. 2. There is no field test available for measuring the density of pipe embedment sand down the sides of a pipe in a trench. 3. It is possible to build into a method spec some feedback that instantly tells the person laying the pipe that the required engineering result has been achieved at all points. (With a performance spec it is usually necessary to wait several days for the results of a few scattered density tests which, as stated above, do not work in pipe embedment sand anyway). 7.1 Discussion 1. Why is it important HOW the required density is achieved in embedment sand? PVC, MSCL and DICL pipes are all flexible pipes. They are not designed to be strong enough to support the trench fill and the future traffic load by themselves. They need a lot of help from the embedment sand around them to prevent them from being squashed out of round by these loads (1). Squashing a flexible pipe out of round puts stresses in its walls that add to the hoop stress from the pressure of the water inside the pipe. Manufacturers usually design flexible pipes assuming that the embedment will be good (stiff) enough to limit the vertical deflection of the pipe to less than about 3% of its diameter under trench fill and traffic loads. If this deflection is exceeded, then the combined stress in the pipe walls from the oval shape and internal pressure will exceed the design stress, and the life of the pipe will be shortened. This means that the embedment sand must not only dense enough to provide the required support for the pipe, but that the density must be achieved in such a way that the pipe is not put out of round by the compaction process. And the most 22 of 52

23 important place for a good density to be achieved is under the haunch of the pipe. This Technical Note was prepared by Ed Collingham, 16/12/2003 (Ex Principal Engineer Geotechnical) Section 8: Embedment Overlay Thickness for Sewer In pipelaying practice, whether water or sewer, the pipe overlay is that part of the pipe embedment which immediately overlies the pipe. The function of the pipe overlay is to act as a mechanical buffer between the pipe and the trench fill above. To act as a successful buffer, the pipe overlay must be: 1. A material that is sufficiently fine and uniform that it cannot itself damage either the pipe or any protective coating on the pipe. For example, for PVC pressure pipes or Greensleeve protected DICL, it is necessary to use sand as the overlay to prevent scratching, whereas for non-pressure PVC sewers, which are less sensitive to minor scratching, screenings can be used. 2. Easy to compact, so that the effort required to compact it does not damage the pipe. Again, sand or screenings are acceptable. 3. Sufficiently thick that any large stones in the trench fill above cannot penetrate through it to the pipe. 4. Sufficiently thick that the compactive effort put into the trench fill above cannot damage the pipe. 5. Sufficiently thick that, even if some were displaced during subsequent construction operations, there would be sufficient thickness left. 6. Sufficiently thick that the minimum thickness achieved is sufficient, even under laying conditions where the control of thickness is difficult, such as in a deep sewer trench. If the above six criteria are applied to normal water main laying, it has been found that an overlay thickness of 150 mm is appropriate. (The overlay material is sand, which is both easy to compact and relatively difficult to displace once in place. The trench is shallow, and so it is easy to control overlay thickness. Quarry rubble is used as trench fill, so there are no large stones in it, and it is relatively easy to compact.) If the above six criteria are applied to normal sewer laying, it has been found that an overlay thickness of 300 mm is appropriate. (Sewer overlay material is screenings which, although easy to compact are also relatively easily displaced. A sewer trench is often deep, and so it is difficult control the thickness of the overlay. Sewer trench fill often contains large stones, and is also often difficult to compact.) 23 of 52

24 In summary, there are at least six factors to be considered when determining an appropriate thickness for sewer overlay. The density to which the trench fill is to be compacted is only one of them. In road reserves we specify 95% of standard maximum dry density for trench fill in order to control surface settlement. In easements we specify 90% of standard maximum dry density because surface settlements are not usually so critical (although they could well be land use, not location, should be the criterion but that is another topic). The difference in the compactive effort required to achieve 90% as opposed 95% would not generally be significant. Indeed, depending on the plant and techniques used, the impact felt by the sewer might well be greater for 90% than for 95%, even for the same trench fill material. To this point may be added the fact that the quality of the trench fill material is more likely to be at the lower end of the allowable range in easements than under roads (ie it may contain large stones and/or be more clayey and so require heavier compaction). It is therefore suggested that there are no sound arguments for reducing the thickness of sewer overlay from 300 mm to 150 mm in easements as suggested in the National Code. This Technical Note was prepared by Ed Collingham, 22/05/2000 (Ex Principal Engineer Geotechnical) Section 9: Embedment Compaction Elastic Modulus VS Density Please be aware that these correlations are incredibly rough! It would have been pushing the limits of the science even to say that the range 3 to 5 MPa implies reasonably good compaction (eg 95% of standard maximum dry density). To differentiate is a flight of fancy, and is more intended to show that they both imply reasonable compaction. I have included the field identification tests I worked out for page B7 of the Blue Book. You may wish to use these tests to supplement the results from the Robert Bock, conical-tip, arboreal, static penetrometer device, with pain-gauge stress-limit control. 24 of 52

25 Table 9.1: Field Identification Tests. Elastic Modulus MPa SPT (blows per 300 mm) (1) Consistency if Sandy Field Identification Test (2) Consistency if Clay Field Identification Test (2) Equivalent % of Standard Maximum Dry Density 5 22 Medium Dense (3) takes a footprint 5 mm deep Very Stiff (4) readily indented with thumbnail Approx 96% (6) 3 15 Medium Dense (3) takes a footprint 8 mm deep Stiff to Very Stiff (5) readily indented with thumb but penetrated only with great effort Approx 93% (6) 1. Based on Hobas design manual correlations. 2. Refer page B7 of Blue Book. 3. SPT range for medium dense is 10 to 30 blows per 300 mm. 4. SPT range for very stiff is 15 to 30 blows per 300 mm. 5. SPT range for stiff is 8 to 15 blows per 300 mm. 6. Correlations not to be used in any manner or for any other purpose than that intended. I have previously prepared spec clauses to avoid just the (very bad) practise of compacting embedment after the placement of the overlay. This practice can clearly only achieve required density in the haunch and side support zones by squashing the pipe. It would be better if they did not compact at all. I have reproduced these clauses below. The pipe side support and overlay material shall be placed and compacted using methods and techniques which: 1. Will ensure that the specified density is achieved uniformly around the pipe, 2. Does not damage the pipe surface or protective coatings, 3. Does not displace the pipe from its laid position, and 4. Does not put any vertical load on the pipe until the pipe side support has been compacted to its specified density. The pipe overlay material shall not be placed until the pipe side support material has been placed and compacted. The pipe side support and overlay material shall be placed in layers, and each layer compacted to a density of not less than 95% of the standard maximum dry density of the material. The layer thickness shall be appropriate to the nature of the material and the compaction techniques used. For pipes 450 mm diameter or less only, and provided it can be demonstrated that all of the foregoing criteria can be met, the bulk of the pipe side support material may be placed in one loose lift to the top of the pipe prior to compaction. 25 of 52

26 I will try to do more research on (a) The use of modulus in specifications, (b) How to test for modulus in in-situ embedment material, and (c) Correlations to relative density. (But if they are going to compact down through the overlay it s all pointless anyway!) This Technical Note was prepared by Ed Collingham, 13/11/2003 (Ex Principal Engineer Geotechnical) Section 10: Controlled Low-Strength Material (CSLM) For Embedments The following provides geotechnical comment on the proposal by the contractor (Leeds) to use CSLM (Controlled Low-Strength Material) for pipe embedment on some sections of Pipelay Packages 1 and 3. CSLM, otherwise known as flowable fill or (in the past) unshrinkable fill, is a sand and cement based backfill material produced in a concrete batching plant and transported in a mixer truck its cost on site is therefore similar to any other concrete product. (1) The main physical characteristics of well-designed CLSM are: It will be free-flowing (with an almost creamy consistency) so that it fills small voids. (Mainly achieved with an air entraining agent and possibly also a plasticiser.) It will have a specific maximum strength when cured eg 1 to 5 MPa. (The usual mix design criteria are that it should be weak enough to allow it to be dug out by hand if required in the future, but high enough to give the required lateral modulus (stiffness) for the pipe and/or vertical/horizontal load carrying capacity.) It will require no additional vibration or compaction after placing. It will have an acceptably low shrinkage during curing. It will reach its design strength in an acceptably short time. (2) The contractor considers that the advantages to him on this project would be: The ability to use a narrower trench (150 mm clearance either side proposed). Avoiding having operatives in the trench (for compaction of embedment etc). (3) Things to be wary of when using CLSM as pipe embedment are: Buoyancy of the pipe in the CLSM which is a fluid with about twice the density of water. (Can minimise by backfilling in several lifts or counter with saddle anchors.) Reduced side support in very weak ground because of the narrower trench. (The side support given to the pipe comes from a combination of the width and modulus (stiffness) of the embedment and the modulus of the natural ground in the trench walls. 26 of 52

27 The time required to achieve the design strength (or, in this application, sufficient strength to enable trench backfilling can be completed) The Leeds CLSM Embedment Trial I inspected the results of the trial on 28 April 2003 (see photos below): The CLSM appeared to be fully cured. (It had been placed on 23 April 2003.) Its strength in-situ was such that it could be slowly fretted away with the toe of a boot or (it was estimated) dug with difficulty using a spade. In terms of its strength in a hand specimen (an engineering geologists method of classification) it could be broken by hand with difficulty indicating a compressive strength of a little over 1 MPa or a very weak (VW) rock. I was informed that: The mix contained 6% cement and had a target strength of 1.5 MPa. One side of the trench was vibrated after pouring, the other was not. It achieved a reasonable strength (sufficient to allow backfilling) within 5 to 7 hours. The supplier considers that CLSM can be produced using 4% cement but not less. Based on my observations of this trial, past experience with the use of CLSM, and general knowledge of its performance, I consider that CLSM would be appropriate for pipe embedment on these contracts. The mix should be similar to that trialled and the points listed in (3) should be considered in the design. 27 of 52

28 A concrete pipe (background) was placed in the trench and the CLSM poured to mid-height. The CLSM was vibrated on the right hand side of the trench but not on the left. The pipe was removed when the CLSM had cured, revealing the quality of the contact. The left side not vibrated. Shows flow banding but otherwise full contact with the pipe. 28 of 52

29 The right side was vibrated. There has been some aggregation of the finely entrained bubbles to give the voids visible in the photograph, but otherwise there is still full contact with the pipe. Vibration is neither necessary nor recommended. The CLSM made full contact with the bottom and sides of the pipe. 29 of 52

30 These lumps were above half height and broke off as the pipe was removed. A lump of the cured CLSM could be broken with difficulty by hand indicating a compressive strength of about 1 MPa. Figure 10.1: The CLSM Embedment Trial Set-Up. This Technical Note was prepared by Ed Collingham, 11/11/2003 (Ex Principal Engineer Geotechnical) Section 11: DS4 Review Difficulty has been experienced with the compaction of several batches of embedment sand on recent pipelaying contracts for the Loveday Irrigation Area, even though the sands complied with the grading specified in the EWS standard specification for packing sand "DS4(a)". It is considered that the compaction problems lie with the very broad grading limits of DS4(a) and PM31. Relevant points are given below and an alternative specification for packing sand is suggested. EWS specification DS4(a) has the same grading as DRT specification PM31. They differ only in that DS4(a) is required to have high electrical resistivity to minimise corrosion. Packing sand for pressure pipes must be a sand - ie not a single size gravel, which could damage the pipe. (DGK) Packing sand needs to be easy to compact so that (i) the pipe can bed into it, (ii) it can be compacted into the haunch area and (iii) to minimise the 30 of 52

31 compaction energy required and therefore possible damage to the pipe or displacement of the pipe in the trench. Packing sand should be free draining and its compaction curve should be relatively insensitive to moisture content, as moisture content is not always easily controlled during pipelaying. (Water entry into the trench; the temptation to moisture-condition by saturating the sand after placing; rain; etc.) The grading curve for DS4(a)/PM31 is shown on the attached (amended) mechanical analysis sheet. Also part of the specification is that the sands be "non-plastic", which implies that the 10% which passes the 75 m sieve should be mostly silt not clay. It can be seen that anything from a clean, single size, 5 mm gravel, to a very fine sand with 10% clay, might pass, despite the vast difference in properties between these two extremes. Also plotted on the sheet are the grading limits for PM61. This is the only other DRT standard specification for a non-plastic sand. It is very tight and would exclude many sands with satisfactory characteristics for packing sand. The single curve is for a sand from a borrow pit in "Maple St". It has a grading similar to those sands which have been difficult to compact. By referring to the bar above the plot (a recommended amendment to the standard mechanical analysis sheet) it can be seen that it is a fine to very fine sand with 9% passing the 75 m sieve (much of which was reported to be clay). Clearly DS4(a) is too loose a specification to guarantee the characteristics required of a good packing sand. It was probably written with the alluvial or glacial sands from metro sand pits in mind - not the much finer, siltier, and lime rich wind blown sands of the mallee country. Nor, possibly, did the original specifiers predict that some metro pit owners would resort to blending fine material with what would otherwise have been an ideal packing sand just to get rid of the fine stuff (and yet still comply with the specification)! The suggested grading limits for a revised "DS4" are plotted on the sheet. The lower curve prevents the material being all 5 mm gravel by forcing some sand into it, while the upper curve avoids it being all fine sand. The "non-plastic" criterion is retained. The suggested alternative specification requires further consideration by Ron Slack (STO Soils, Materials Sciences Centre) to ensure that it is appropriate - in particular that the grading is not so tight as to exclude materials which would have satisfactory properties, and also that the 10% passing 75 m criterion is sufficiently strict by itself to ensure that the material will be free draining. 31 of 52

32 Section 12: Pipe Laying Groundwater Control Specification Clauses GROUNDWATER CONTROL The Contractor shall remove any water that may enter or be found in the pipe trench or associated excavation while the pipes are being laid, and while any other works under the contract are being constructed. Where the trench floor is wet but the trench is not actively making water, the SR may allow a stable working surface to be created by over-excavation and the placement of a bed of 20 mm aggregate wrapped in geotextile. Where the trench is making water slowly, the SR may allow the trench to be drained using a bed of aggregate as above, draining to a pump sump in the trench. Drainage to the sump may be assisted if necessary by an agricultural drainage pipe laid in the screenings. Where the trench would be expected to make water rapidly, or is found to make water rapidly, a wellpoint dewatering system shall be installed and the watertable lowered to below the floor of the trench before the commencement of pipelaying. The wellpoint system shall be operated until such time as there is no danger of flotation of the newly laid pipes and the trench has been backfilled to not less than 150 mm above normal groundwater level. Where approved by the SR, costs of materials for stabilisation and drainage will be reimbursed by the Principal. Water from excavations and wellpoints shall be disposed of in a manner that will not contravene EPA regulations, cause injury to persons, or cause damage to property, the work in progress or completed work. The Contractor shall be responsible for the consequences of, shall bear the costs of, and shall settle any claims for, personal injury, death, damage to property, and breaches of EPA regulations, caused by the dewatering operations. This Technical Note was prepared by Ed Collingham, 06/06/2001 (Ex Principal Engineer Geotechnical) Section 13: Pipe Trench Width Discussion WATER SUPPLY AND SEWER CONSTRUCTION MANUALS (The Explanation of Differing Trench Widths Specified in the Two Manuals) A 600 mm minimum trench width is specified for sewers (150, 225 and 300 mm), as sewers are generally deep, and there is a need to ensure reasonable access for personnel during placement and compaction of the embedment. A narrower trench width is allowed for water mains, as mains are generally shallow, and therefore reasonable access can be gained for 32 of 52

33 placement and compaction of the embedment within a narrower trench than is required for the same size of sewer, particularly for the smaller diameter pipes. No maximum trench width is specified for sewers, as sewer pipes in the 150 to 300 mm size range (with the specified embedment) are strong enough to support all trench fill and traffic loads without relying on the shedding of some of those loads to the trench wall. Not specifying a maximum width allows the contractor to use battered or benched cut excavation if these are more economical. A maximum trench width is specified for water mains as it is desirable to disturb the ground for the minimum width possible, because anchors and thrust blocks must be bear on undisturbed ground. Sewers do not need anchors. Specifying a maximum trench width for a water main does not unduly restrict the contractor, as the mains are shallow and he is unlikely to want to use wide battered cuts anyway. Having specified a maximum width for the above reason, it is taken into consideration in selecting the "support type" (H1) for the design of concrete pipes. This Technical Note was prepared by Ed Collingham, 09/06/2000 (Ex Principal Engineer Geotechnical) Section 14: Pipe Laying Groundwater Control Using Screenings The groundwater appears to be related to the ornamental lake alongside the road. Judging by the greenness of the vegetation (roadside grass and trees even on the other side of the road) this lake is holding up the watertable over quite an area. Uphill of the lake all the vegetation is dry and brown. The watertable is also being held up by the blockage of the watercourse downstream of the culvert under the road. The soil appears to be stiff yellow clay that is not allowing water through because it is fissured. It would be necessary to get the groundwater under control and ensure a dry trench floor before laying the pipe. Whatever approach is taken to do this, should also ensure that there is no erosion of the pipe embedment or trench walls/floor after construction. I suggest you consider the following steps: 1. First clean out the watercourse downstream of the culvert so that the ponded water upstream of the culvert can drain away. 33 of 52

34 2. Open up the full length of the trench from the present point to the culvert so that seepage can drain out through the culvert (at least to culvert invert level). 3. Check that the soil in the trench walls and floor are stable even though water is seeping from them - ie they are not soft clay or loose sand. 4. Judge the rate of flow of seepage down the trench. 5. If it is considered that the seepage flow is low enough to be carried down a bed of screenings, then judge what depth and size of screenings would be required. (See Clause 4.5 in DS9.) For example, if it is decided that a 100 mm thick bed of 10-7 mm screenings (normal SA Water approved PM43 sewer screenings) would handle the seepage, then over-excavate the trench floor by 100 mm, place geotextile on the floor and up the sides, fill to a depth of 100 mm with the screenings and wrap the geotextile over the screenings. The pipe may then be laid with DS4b embedment in the normal manner. Or if it is decided that more capacity is needed in the screenings, then consider using say a 150 mm thickness of mm screenings (PM41), and proceed in a similar manner. Or if it is decided that there is too much water flow to be practically handled by screenings then it may be necessary to consider wellpoint dewatering. 6. It will probably be necessary to dig a sump near the culvert and pump from it to control seepage there. Be careful of the detail here. Nowhere should sand be in contact with screenings or screenings in contact with the trench floor or walls - all screenings should be wrapped in geotextile, and all water should be drained or pumped from the screenings in a way that ensures the screenings cannot be lost. Repeat the same approach the other side of the culvert as and if necessary. I would be interested to know how it all goes. This Technical Note was prepared by Ed Collingham, 14/02/2001 (Ex Principal Engineer Geotechnical) Section 15: Pipe Laying in Embankments Reclaimed Water Main through the Lagoon Embankment The section of the reclaimed water main through the embankment of Lagoon #3 shall be embedded in high-slump concrete with a maximum aggregate size of 10 mm to mid-height of the pipe. Precautions shall be taken to prevent flotation of the pipe in the concrete eg by filling the pipe with water, loading it with sandbags and/or strutting. The remainder of the pipe embedment, from the top of the concrete to 150 mm above the top of the pipe, shall consist of well-graded low-plasticity sand to 34 of 52