Scribd Logo
Upload

Welcome to Scribd!

  • 318 views

    Hydraulic Calculation

    Uploaded by

    Bader Shrbaji
    Hydraulic calculation for GRP pipes

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd

    Hydraulic Calculation

    Uploaded by

    Bader Shrbaji
    318 views46 pages
    Hydraulic calculation for GRP pipes

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Hydraulic calculation for GRP pipes

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Download as pdf
    318 views46 pages

    Hydraulic Calculation

    Uploaded by

    Bader Shrbaji
    Hydraulic calculation for GRP pipes

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Download as pdf
    of 46
    Hydraulics Characteristics Of AFIL’s FRP PIPE Amiantit FIBERGLASS Ind. Ltd. (AFIL) Dammam- Saudi Arabia 1.Hydraulic Characteristics Summary of AFIL’s FRP Pipes 11 Introduction 1.2 ERP Pipe Roughness 1.3 Recommended Velocities 1.4 Internal Diameters 1.5 Conelusion and AFIL’s Capability 2 Attachment I: SINTEF Report on Pipe Roughness in GRP pipes 3 Attachment 2: Abstracts from AWWA C-950 Manual 4 Attachment 3: Certificate From SWCC of Saudi Arabia 5 Attachment 4: ‘Typical Head Loss Calculations Yoo Characteress Pa Hydraulic Characteristics Of AFIL’s FRP Pipe 1.1 INTRODUCTION Amiantit FIBERGLASS Ind. Ltd. (AFIL) produces FRP pipes by continuous filament winding machines, by reproducible processes. All these pipes are provided with resin rich interior layers, providing very smooth inner surfaces. This smooth interior surfaces results in very low fluid resistance, thus increasing the discharge of the fluids for same diameter of traditional non-FRP material. 1.2. FRP PIPE ROUGHNESS For hydraulic analysis of the every piping system, Pipe roughness is of the concern, One of the FAQ by the Hydraulic Engineers / Consultants / Contractors / Clients is what is value of FRP pipe roughness, This Roughness is being used in various forms in various equations of hydraulic analysis. Find below the summary of the mean values of Roughness Parameter of various equations. ‘This is the latest value based on the experimental studies. Roughness Parameters (Mean Values) Flow Rate ‘Cole Brook—White | Manning | Hazen — Williams | (m*/ Br) € Or K(mm)_| M(m!*/s) | C10* m7) ~~ 1410-2860 0,029 104 __146 ‘Above values are based on the experimental studies carried out by Owens Coming, and SINTEFF from Norway. For easy reference a copy of the report is aitached (Sce Attachment 1) In fact AWWA C- 950 also recommends for the usage of similar values. Find attached the abstracts from the appendix of AWWA C- 950 in attachment 2, which confirms above values are in good agreement even with the international standards, LANTIT Apart from above, the interior pipe surfaces, typically remains smooth over time, in most fluid services. Therefore fluid resistance will not increase with age. This has been demonstrated, when few FRP pipes under operation over the decade were inspected and evaluated. Such one certificate is attached for easy reference in attachment 3, Based on the independent reports and evaluation from the site it can be safely concluded that the usage o f these hydraulic Roughness parameters should suffice the requirements. Attachment 4 will provide the typical information about the head loss calculations based on the Cole Brook White Equations both in implicit and explicit form. 1.3 RECOMMENDED VELOCITIES For AFIL's FRP piping it is recommended to have a velocity of 3 m / sec. Higher velocities are also possible, please consult AFIL for further advice. 1.4 INTERNAL DIAMETERS, In general AFIL's FRP pipes have higher internal diameters. Actual inside ameter is almost same as nominal diameters. Therefore, it is recommended to nominal diameters as internal diameters for analysis purposes, 1.5 CONCLUSIONS and AFIL’s Capability Hydraulic analysis can be carried out based on above information coupled with appropriate project requirements and considering equivalent lengths for the various fittings and equipments, Moreover AFIL is equipped with software PIPENET from UK to carry out the Hydraulic Analysis including pipe re-sizing and surge analysis, NV ZAMIANTIT ATTACHMENT # 1 AFIL Design Department STF22 A96309 Gradering: Unrestricted Pipe roughness in a glassfiber reinforced plastic (GRP) pipe SINWEF Civil ond Environmental Engineering SINTER SINTEF Civil and Environmental Engineering Water and Waste Water Address: N-7034 Teendhein RoAWAY Leeation; Kisbuveien 188 Telephone: +47 73.5929 00 Fax $4773 892978 Tolox: 88 620sinin Entorprise Ne.: 848007028 PIPE ROUGHNESS IN A GLASSFIBER REINFORCED PLASTIC (GRP) PIPE amy Odd Ade Tveit == — + Owens-Coming Pipe Technolgy erORTNO, jecsericaron RF STF22 A96309 |Uneswicted [Roger Josefsen fea. Fron PAGE —fsaN PRORECT Wo Sor GERI Unrestricted | 82-595.9809-4_|605323.00 2p+2e ELECTRONIC FLE CODE JosceuNaay RESORT ~ rmenonrenearornseonst [Arve Berg [rtecooe are }605323.00 1996-06-28 ROLES, cy Bjthar Elkebrokk Dept. Manager jaasrancr ‘Owens-Corning Pipe Technology wants to monitor the intemal pipe roughness in a 700 mm glass fiber reinforced plastics (GRP) pipe over a period of time, SINTEF Civil and Environmental Engineering, Department of Water and Waste Water has planned and will also carry out a monitoring programme for measurement of pipe roughness. Pipe roughness has been determined by measuring flow rate and head Joss for a 348 m long pipe section. Based on these measurements pipe roughness coefficients are calculated, Hydroteknikk Hydrotechnology | concur Flow Sugmning anaceonvarwons) | Friction Faiksjon | Roughness Ruhet | Glass fiber reinforced plasties GuP © sumer 2 ‘TABLE OF CONTENTS ABSTRACT... 3 PREFACE ...... Denes 4 1 INTRODUCTION .. « ae - wae 5S 2 MEASUREMENTS . 21 Theory 2.1 Colebrook-White 21.2 Manning 21.3 Hazen-Williams .. 2.2 ‘Measurements, pipe system 23° Measuring devices ....... 3 RESULTS ......... 3.1 Headlods 6... . 3.2 Pipe roughness «+... 6.ee ee 3.2.1 Result summary ..... 32.2 Colebrook-White . 32.3 Manning... 6++ 324 Hazen-Williams .......+. 33 Accuracy in measurements and calculations .. 4 CONCLUSIONS AND FUTURE WORK «2.0.6... 0. 000s ees 5 REFERENCES ... (OAT 05323-bw5tL . Owens-Coming Pipe ‘Technology wants to monitor the internal pipe roughness in a 700 mm slassfiber reinforced plastics (GRP) pipe over a period of time. SINTEF Civil end Environmental Engineering, Department of Water and Waste Water has planned and will also carry out a monitoring programme for measurement of pipe roughness, Pipe roughness has been determined by rheasuring flow rate and head loss for a 348 m long pipe section. Based on these measurements pipe roughness parameters are calculated. Calculations of the roughness parameters, measured at inital conditions, are based on three different flow rates. The result summary is shown in the following table. ‘Roughness parameters (mean values) Flow rate Colebrook ‘Manning ‘Hazen Williarns tm) k.(mm) M (m"? js) C0" m/s) 1410 - 2860 0.029 (104) 146 out of range for the Manning formula OAT\R605525-b 6:61 PREFACE, This report summarizes the procedures and results of pipe roughness calculations based on flowrate and headloss measurements on a 700 mm glassfiber reinforced plastics pipe produced by Owens-Corning Pipe Technology. ‘The measurements were taken on 1994-12-08, 1995-05-19 and 1996-01-22 at the Gollfjellet water supply scheme in Bergen. The work was carried out by Halvard Gundersen and Odd Atle Tveit, SEVTEF Civil and Environmental Engineering, assisted by Gudmund Bruris, the Municipality of Bergen. OA TINEOS323- WS NOL (5) Sibir 5 1 INTRODUCTION ‘Owens-Coming Pipe Technology (Veroc) produces glassfiber reinforced plastics (GRP) pipes for a wide variety of applications. In the context of water supply systems, GRP pipes are often used as trunk mains, In the city of Bergen a new source at Gullfjelet has recently been connected to the municipal water supply system, The Gullfjellet scheme includes a 700 mm GRP pipe section from the intake at lake Svartavatnet to the Silhall water treatment plant. ‘The objecitve of this study is to determine the change in pipe roughness as a function of time. Calculations of the pipe roughness are to be based on measurements along a 348 m section of the pipe. SINTEF Civil and Environmental Engineering (former The Norwegian Hydrotechnical Laboratory) has been assigned a contract to plan and execute a monitoring programme to determine changes in pipe roughness. ‘The scheme construction work was finished and the municipality of Bergen started to draw water from Svartavatnet early spring 1994, The measurements which are described in this report were taken on 1994-12-08, 1995-05-19 and on 1996-01-22. (OAT 65123-86861 © sane s 2 MEASUREMENTS Pipe roughness can be calculated by measuring of head loss and flow along a given pipe length. Total head loss is the sum of friction loss along the pipe length, including losses due to pipe joints end bends: ay +b, @) Total head loss was measured along a straight line pips section, 348 m long made of 29 sections of 12 m, Therefore there are no bend losses. Losses due to pipe joints are assumed to be minor. In the calculations these losses are included in the total friction loss along the pipe. ‘The friction loss can be calculated using several different formulas. In this report friction loss has been calculated with the following formulas (Beyum et. al, 1980): + Colebrook-White + Manning + Hazen-Williams 21.1 Colebrook-White ‘The Darcy-Weisbach formula: (m) - 22) where: = friction loss (m) = friction coefficient (friction factor) length of pipe (rm) Flow (m/s), internal pipe diameter (m) = gravity constant (9.81 m/s4) euorms OAT E05323-bow6 NtbL . gux Pipe roughness can be described by the Colebrook-White roughness coeff - 231) a.nd Re VE | where: Re = —49 Deny (Reynolds Number) y= Kinematic viscosity (m/s) 2.1.2 Manning Manning's formula: O=M- a+ RY. Tt where: = flow (ms) = Mannings numiber (feiction factor) = pipe cross section area (m") = hydraulic radius (m) = A/P where P = wet perimeter (m) = slope of hydraulic grade line (r/m) Hw> ZO ‘The manning formula is valid within the given ranges (Pedersen 1988): a.7< 8 < 300 x K<10 where: (OATRNG05323-bow6 11 (3) 24) 4.1) 2.1.3 Hozen-Williams ‘The Hazen-Williams formula: Q= 6,67 +c DHE» r08 >) where: Q flow (Us) C = Hazen-Williams C (friction factor) D ipe diameter (m) 1 slope of hydraulic grade line (%e) 22 Measurements, pipe system Pressure head measurements were taken at both ends of the GRP pipe, on an uncovered section in the wanel downstream of the source at Svartavatnet. A Differential Pressure cell was con- nected 10 the existing couplings in both ends of the pipe section via « 390 m long 1/2" PE pipe. Flow rete was measufed simultaneously at Silhall, about 1500 m further downstream of the tunnel, See Figore 2.1. © SVARTAVATN D=652.15mm L=1800m | 348m | ans _ ~360 ~360 ~s40 SILHALL ay = . 300 © Figure 2.1 Longitudinal section Svartavata - Sihall. OiVR605329-bawi Nt © sinuer 9 Measurements were taken at noflow condition, i.e. the system was shut off, and at three: different flowrates. The flow rate was controlled by a valve located downstream of the flow meter at Sithall, Pressure heads were logged on a computer at Svartavain while flow rate was read manually off the flow meter display at Sithall 23 Measuring devices ‘The differential pressure was logged on the computer with one second intervals using a Fuji FCX Differential Pressure Transmitter (see Appendix A). Mean values per minute were calculated and stored by the computer. + Range 0 - 3.2 mof water gauge + Accuracy <2 0.1 % of span (3.2 mm) At Silhall, flow rate was meesured with a 600 mm electromagnetic flowmeter, Techmag DI652, from Endres + Hauscr (see Appendix B). The flowmeter has the following perfor- mance data: + Range 0 - 6000 m’m + Accuracy #2 0,5 % of reading when flow is within 20 - 100 9% of measuring range OAT R605323-box6 NEBL © sie 2 3 RESULTS: 31 Headloss Table 3.1 shows the average head loss for the respective flow rates for the three different measuring dates. As an example, minute values for the measurements on 1995-05-19 are shown in Appendix C. Table 3.1 Average head loss for different flows and measuring days, [Date of Flow rate | Average value measurement (n/n) head loss Flow fate Ga) (om water gauge) 1400 | 2100 | 2800 1994-12-08 1430 0390 1907 1994-12-08 2860 L410 Lr 1995-05-19 1410 0.405 2.037 1995-05-19 2103 0.845 1911 1995-05-19 2844 1473 1.821 1996-01-22 1416 0373 1.859 1996-01-22 2120 0.789 1735 {1996-01-22 2196 1327 1.697 For a constant flow head loss can be expressed by: AH=r-@ where: r © constant factor 32 — Pipe roughness Pipe roughness, including losses due to pipe joints, can then be calculated using the Cole- brook-White, Manning and Hazen-Williams formas. The following data are used: D = 692.15 mm L = 348m y= 157+ 10-Smi/s_ (Water temperature 4 *C, May) 152-10 m¥s (Water temperature 5 °C, December/January) OAT RN6OS325-bsW5 N61 © sues 32.1 Result summary ‘The results are shown in the following table, Table 3.2 Roughness parametess. 1 Friction parameters (mean values) 7 Flow rate —— (ns) Colebrook Manning Hazen-Williams (mim) M (m' Js) C (10-1.38 m/s) 1410-2860 | 0.029 (04) 146 * ut of range for the Manning formuta 32.2 Colebrook-White ‘The roughness coefficient is given by Eq, (2.3), Calculated roughness coefficient kt for the different flows and measuring days are shown in Table 3.3. Table 3.3 Calculated roughness coefficient, Dale amar ata wwe] wet Fe im ousessa| 1400] oer tos] 0200] aorasa | aateves| 01a) cbdee.o4 | zas0 | Teed ait | 1410] 00123 | asaevos| 0.025 | 19-msi9s wie [sen 10s] oans| _oowsr| 4.5a€+05| 0.088] semaves | 20s se42 155| oaae|aotser| eases] 008s) remaies | __ 2008 re00 zo] 17a] _ooizes | szscvos| 0.0 zajana6 | tae | 9038 sos] oars] oorsso | a7eevos| 0.00 [7 zajenas | 2120 | sae ss7| aves] _aovass | 7.taev0s) 0.00) 2ajenae | __2706 | 7767 or] taer| _ourzia| e4oesos| —_o.or Average] aed Sid. eu] 0.028] Dev. (| 85.8 (OATVR\605323-bs96 BE © susirer 2 cor owe meas ee Bl] etemsies oe zeros | 0030 lat 2020 as ae <0 L mi io wie 0b aad ares aeee aso wh Figure 3.1, Calculated roughness coefficient versus flow rate aro 060 ‘030 2 0.080 .0%0 0.020 e010 2.000 ais yas a 1 Pesan 96 Figure 3.2, Calculated roughness coefficient versus measuring date. Records from the municipal water treatment plant show a water temperature of about 4°C in May and about 5°C in December/lanuary. The kinematic viscosity, used in the Reynolds number, is based on these temperatures, ‘The calculated k- value based on the different measurements shows a standard deviation of 85 %. The k - value is low and is.very sensitive to variation in the measured head loss and flow rate, Table 3.4 shows that a deviation of the roughness coefficient of +85 % represents a deviation of flow in the order of 1.6 % to 2.4 % for the given flow range. OA VE905323-b wb NAL SINUEF Table 3.4 Deviation of flow rate due to deviation of roughness coefficient 43 (mm) ‘An (m) Q (may 0.029 85 % 037 £30 % 1400 1.6 % 0.804 4.0 % 2100 2.1 % 1.36 4 4.7 % 2800 £24 % Based on the measurements, k = 0.029 mim is the mean k-value, The above table shows that the flow, Q = 1400 min, will only decrease by 1.6 % if k is increased by 85 %, while head loss will increase by 3.0 %. 32.3 Manning ‘The Manning number is given by Eq, (2.4). The Manning number M calculated for the different flows and measuring dates are shown in Table 3.5, Table 3.5. Calculated Manning number. oa cram ama] wor a i 08-des.94 1430| 397.2 1.08 *"0.339| 1017 edes3i azo Teka zat sare ‘038 omsias ae 04 es ca "a-ms:95 584.2 ass| aaa 3018 ‘omnes ze 700 2.10 tar] to40 [pase aie asaa[ 108 ears] 1000 zalan96 2a sek 157 are 1060 Zeist are 7787 207 20H tore ~ Averagel 403.7 soy rf ‘ev a4 OATVR\605323-box6INibL 108 Cor 108 B19 msi 95 1 7259096 = 100 22 rs o 8 tio 18 4430—=«7103-=«2z)7O8 eA O Fw (nn) Figure 3.3, “Calculated Manning number versus flow rate. 109 108 onset teaey5s Patan Figure 3. Calculated Manning number versus measuring date. “The calculated Manning numbers show 2 standard deviation of 43.2. The calculated flow will have the same deviation, as the flow is linear proportional to the Manning number, (Eq. 2.4). “The validity of the Manning formula was checked using Eq. (24.1): 073 0,025 - 107 6920 > 300 ‘This shows that due to the smoothness of the pipe (small roughness coefficient k) measuring conditions are out of range for the Manning formula, Therefore the Manning formula should not be further used for this GRP pipe, OAT\R605323-ban6BE aa © SIRWEF 32.4 Hazen-Williams ‘The Hazen-Williams coefficient is given by Eq, (2.5). The Hazen Williams coefficient C calculated for the different flows and measuring dates are shown in Table 3.6. Table 3.6 Calculated Hazen-Williams coefficient. Date (mam) CT) Mimis)| htm] 7 | aed ‘eeo] sa | vee | ost anal a eden zato|——roxa| a0 408|_ ara ‘Beis “aia aon | 100 tae] ven 19-395 2103 5842 455) 0.845] 2.43] 1428 — ‘ul 7ma| a0 [sara ao] iano Wis] sa a0 [ora tor] as ziao[ tena | a] areo] a7] waa zi|__vrar [zor] saarl aan] wo or Bavey[ at Berta) 2 “The calculated Hazen-Williams coefficients show a standard deviation of £3.1. The calculated flow will have the same deviation as the flow is linear proportional to the Hazen-Williams coefficient (Bq. 2.5). “0 _ aaa 148 siemaiss |_| ie . Bh | ctzimse 7 bie -| ee p : | | a ' ' oli Prana sear rr “ Pw a) Figure 3.5, Calculated Hazen-Williams coefficient versus flow rate. OnirvR605323-boe61t 36 i418 21202798, 1416 2860 am) ovetue Figure 3.6. Calculated Hiazen-Williams coefficient versus measuring date, 33. Accuracy in measurements and calculations Based on the given accuracy of the measuring devices it is possible to find the theoretical accuracy of the calculated friction factor, Hazen-Williams C. “The measurements on 1995-05-19 show 2 head loss of AH = 0.405 m at a flow rate of Q= 1410 mf, From Eq. 25: Q=6.67-C- D0 ‘The accuracy of the friction factor e, is determined by the error of the flow eq , the error of the slope of the gradeline ¢; = qq, and the error of the diameter ep. ‘Accuracy in flow rate measurements, according to the manufacturer, is (see Section 2.3): ig = £05 % Accuracy in pressure measurements, Coleman et. al 1989 (see also Section 2.3). £32 mm = 22-100 % = £08 % 405 ‘OAT VRG05323.b6 tot Accuracy in diameter of the pipe is given by the factory to: zy = 20,25 mm = 225 . 100 % = + 0.04 % 62.15 Based on the given accuracy of the measuring devices, the theoretical accuracy in caleulation of the Hazen-Williams Cis found by: 2 feBe054 620.00 of = 2/00 SP ORDO OF STEMI ~ 20.7 ‘As can be seen from these calculations, expected errors are small, The error in flow rate rmeesurement and head loss are the dominating errors while the error in diameter can be neglected, Data from the pressure measurements (Appendix C) show the following variations: Table 3.7 Maximum deviation in measured pressure, Flow (m'/h) 0 1410 2103) 2844 Deviations (mm) 0 2.6 ii u “This shows that there are small variations in pressure in the pipe and that the variations increase as the flow increases. These variations are real variations of the pressure in the pipe and have nothing to do with the accuracy of the measurements. When the pressure is fully constant (zero flow) then the DP cell also shows a fully constant pressure (Appendix B). Based on the measurements, the calculated standard deviation of the Hazen Williams friction factor is 43.1 %, while the expected accuracy based on the performance data of the measuring, instruments is +0.7 %. ‘The Differential Pressure Transmitter was tested in the laboratory, both before and after the field measucements, This control eliminates that the instrument can be the cause for these deviations. ‘The flow meter can not be fully tested on site, Normally these devices show an accuracy within the given ranges when they are installed properly. OAT IN6DSI23-boo6 NEBL © sine 1 During the measuring period 1994-12-08 to 1996-01-22 the flow meter has been checked and adjusted twice by the dealer of the equipment, This happened between the measuring dates 1994-12-08 and 1995-05-19 and between the measuring dates 1995-05-19 end 1996-01-22. ‘These minor adjustments on the equipment might therefore be the cause of the above mentioned deviations in calculated roughness. The deviations can not be explained by change in pipe roughness only, since the measured losses arc highest at 1995-05-19. “The water Jevel in the source Svartavatn (and then also the intecnal pressure in the pipe), was 08 m higher in December than in May. ‘Theoretically, the pipe diameter increases with in- creasing internal pressure. In this case this phenomenon can be neglected. An increase of 0.8 m in pressure increases the pipe diameter by 1.6 - 10 mm, A significant difference in water temperature may also have an effect on the pipe diameter. However, the records from the municipal water treatment plant show a water temperature around 4 to 5°C on the three measuring dates. ‘OAT R605323-b wéINtbL © sinner . ws 4 CONCLUSIONS AND FUTURE WORK ‘As can be seen from Chapter 3.3 the calculated roughness parameters show that the pipe is very smooth. The different measurements show a deviation in roughness parameter which would atuibute to @ deviation in flow of 1 - 3 %. ‘The most critical factor of the measuring system is the flow meter. Minor adjustments on the instrument may cause small changes to the measured flow. This subsequently will cause deviations in calculated friction factors. Based on information from the Municipality of Bergen it is not likely that the flow meter will be adjusted in near future, However, this should be verified before new measurements are teken, Future measurements will give a verification to the standard deviation of the different roughness parameters. Through future measurements it will also be possible to see if there is any change in the roughness parameters as a function of time. ‘The measurements of 1995-05-19 show a slightly higher head loss compared to the measure- ments of 1994-12-08 and 1996-01-11. The DP cell was checked both before and after the measurements were taken, The deviations are therefore most probably explained by a small change in the flow rate measurement system, OAT G05323-bw 61 (©) SUTEE 20 s REFERENCES Boyum, A., ctal. 1980: Sanitary Engineering. Basic course, patt I. Department of Hydraulics and Environmental Engineering, ‘The Norwegian Institute of Technology. Coleman, H.W., etal. 1989; Experimentation and Uncestainty Analysis for Engineers. _ ‘A Wiley-Interscience publication, ISBN 0-471-63517-0. Pedersen, FLBo 1988: Hyélraulies for Civil Engineers. DTH, Lyngby. ISBN 87-7381-046-0. ONTYRNEO5323-bs06 OATYR 05 325-bw6 Neb APPENDIX A aa FC -series Elecironic Transmitter SPECIFICATIONS seuss (Model FHC/FKC) dle Rongee 2Mesel Osa, 6/60mbar 0... B4°7E40Mbor + O....373008 0, O13/1.Sbor —(0.,..2/20baN9 On ‘Accuracy Rating at calibrated span "The accuracy ls puarantoed forall app ‘eccuraey is guarantaed for al application rang Uneasy 0.05% Tempemusene: | 12 osnssere o2surserc ” at max. span Total 0.8%/55°C OS%/SS°C_ a deroatmax.spon 02 %/10, 32 bor or 100 er : Stove Pressure Eftect — 2oatmmeapen _O2/10.820e __avvicoser ast Span O2/82 bar 7/100 ber oF Over Pressure Effect on Zero at max. span O3%/0, 82 bor o.3%/1€0bsr 08%/420b0r 075% Output, Power Supply Zero Elevation/ Suppression rice Turndown Ratio Elecuieal Damping forcet “10 Temperature Umit Pe for Amp. Pressure Rating setae 10,420 Overrange Limit Upwrn Chesstication of Enclosure Explosion-proot fs Intinsc Safety a Wetted Part Moteria ares Process Connection . 1/8-18NPT weight toe hia 20%“ Opean °¥ rsure Rung 4203 ony“ Bnet cine he rv fo 1/1009 te meen IFEX 125 Sut Tae eo @sime OAT WRBOSI23.-b6IbE APPENDIX B Teknisk information Magnetist-induktiv flowmater Introduktion PULSMAG V, en del a! VARIOWAG- familien,ucteres | mange stertelser og Hengatiaiviaiager VARIGIAG-flovimalerne (4s med man- (ge meksaiske tlslutningsmuligheder {litines, waler type, DIN og ANS! Hanger ets) ra nominel diameter DN 3 til 1000 en (0 tH 28600 m’in) og har Semme mikraprocessorbaserede eiak tron Instrumenterne bruges il méting af vand, splidevand, syrer, lud, slam, Ki ster et., der har enledringsevne storre fend5 pSlem, De store fordeieligge iden femme opsta‘t- og betjeningspro¢ ure, sbvel som stor drilsikkerned og Tangiidestebiitet. Maleprincip Malepdincippet er baseret pa Faraday’s Inaveerst speending magnelisk induktlon alsland meliem elektrader (izengee af leder) vy shastighed af leder (medie) en magnetisk-nduktiv flowméler re> prinsenteror det elektrisk ledende edie (min. 5 yStem) en leder, der tores gennem et magnetfelt. Den Inducerer fn speending, vinkelret pi flowretnin ‘gen. Sprendingen er proportional med ‘den gennamsnitlige flowhastighed. Spamndingen bearbejdes | en hol impedant, low nolse forstaecker og om soatles til et skalérbart, analogl ude gangssignal (04.20 mA) 09 et digitalt signal (impulserivolumenenhed). {PULSMAGV bliver magnetfeltei gene- toret ved en positv og negativ strom (OC-system), Alle tunktioner er miko processorsiyrel. Det modforer bl.a., al man kke skal Korrigere for uipunktsta billtel og nulpunktsait PULSHAG V LSversion (adskil elekironik, max, 200 mi) Alle vigtige funktione: i maleren er frit programmerbare ingen for de gine Speciation ‘maleownrad strom, puls 09 Irekwensudgang, tlds onstant, Keybestromsunderteytkelse, interlerens alblzencing. Udgangssig nalerne er selfoigelig galvanisk ad- kilt Ira indgangssignatezne. Matesystem PULSMAG V bestir tet mélerer 01652 (99 en mikraprocessorsiyrat forstazrkes toned HI 6520, Ved slandard-eppilkationer anvendes kompakienheden, tvor sensor+ trans: mittererén anhed. Ved hoje temporat: fer(>130°O), Exzone og alter kundeon: ke levores adskit version (Sversion, max, 200m Valg af dimension Den bedst egnede cimension kan fn es vina, aomogramme:, fio. 2,09 te bollan, fig. 1. Mileornrdds kan vatiges feliam 0,89 0g 28500 nth, Malorne Kae Ubreces ipravestand. Flowhastigheden skal, s8 vidt muligt, alpasses med mediels fysiske egen skaber. For medies, det nar tendens til at allojte | roret, teks, splidevand og slam,vealgermenennastighed >2mis. Floviastighedenoges vedreduktion al sensordiameter, For slidende medisr, teks. katkeealk, ligger den ideelle flowhastighed und 2s. Til slidende medier anbetales lo ting | blodgumm Vad vancige oplosninger opnas der er faringsmapssigt de bedste malerasul: later med en max, flownactighed p& 1.10 ns Dimension; méleomrade DY [P| Adleomede Vi mm, toon | omasnn mar ais o35 355 [« B 2 ow & | % 0 0 1 e 6 & © ‘ 4 0 ° 0 4 Q a a a, o 0 0 0 4 Q 0 0 0 a Q Ml ® 8 6 me 08 w x Fy Fy 9 Tokniske data olameter Baiteye Maleror, mate Flangemateriate ateroretoring Elekivodemal Max, tadelige bnedislema Maxcemsuem, Min, tedaingseone Nalecrnrad Netspaxnging Erfetiorug Beseytctees Indgangs: oasis adgange: Siren Tiaskonstant Stenauingte- dgang Nojagtighed ht gp yowoe 2 Medietemperatur ‘Ging teimporator rsyings spamnaleg opvacraningsta inaeo Usies Maleteknix Industriel autor 15.1000 ta 1) PHEOON 60 2B em BNIGON 25” 200mm NOON 2501200 mm S'S. 4903 Als Sos) Sti, for moctange 6: ter sin 290 VR seean 18. 600 PTFE 0825-10006, 8's oll srs (arst 318TH, Haselloy, Haragum. O..¢ e0%e Slaggum ~20""4 120% PTFE a § 20° Pio. +sorefagereg au seta deastiozzov a-aste, Solo 5% atv0c 40-15% Ga nevacraw 1 6541967 09 P 60 som ptien) Brera, foxmstebt av (CaterelSttptage > 200 p1ptlanm(> 100 Gohny 04.20, program ‘eibar, PL «780 ohm 05.106, programmer bi 24,20 msiorogram imessan.imex 2H, soma programmer. Say eer dbon eltenor fae 20 He rogram mera! S20mA sacabetingelser ht. OIN 12200 ‘ Be matisering ven lagi lasttliow- ho anatye rising EnaveasgHavaer a8 Poaeelgdare 102 ebony fnaigravaz The 2380 Fas.31 679045 Dimensioner 20 15 9 20(PN 49, O28. 200( 16) 2801600 0} Malenojagtiched ih.t {lorencobellngatser: Snslog vayama fox, malesmeace: ome ‘21% af ante! inp 20-00% Fe al mas iaozon ft aktuel ‘0% ost ais option Soar a axel z '20-100% Aepetersines 30, a mas. Fett sendringerlorvehotdes vedi vera on Ala yoy oo wt | gy |e | a s6| a | | S| pm | ae] alos S00 | ve | 22 | da | Sos] 1s cam |" | Se | fe | | tz Fa ‘w]e a] mee io yo| 45 | 2] 2s 3 S| S| ee | dsaa i S| | ro] a0 2 Go| a | 1] Sieo sao | 10 | Smo| coe] too | Sue io aaa me] 912) Soon 7 Seo] 903 | sae | ses 2 ian | ras | 19 | Stes ‘ fate | Hes | 2a] ens 10 190 | ais | et | eae Endress-+Hauser sewe ER SINTIEH OAT UR G05323-LENKbI APPENDIX C 23 Sheet pare | T far = ornaih at 112500 fs fime__[DP Pahmmes [tw a) [0 112500 112600 = 4410 math |_112700) 77800) 0008 OE 712800) gz anne 1 i tt | mI 713000 £ ome A 1ST { T1301 B cae PAF =e 773200) Bor T 713300] * rt t 1713400 ah ed “esootww7oo”—vvesea=—«wvStOO «00 «118800 175600) = T7709) [713800] [ 713000 4000] 0.464] Q=2103 mah TraT0o] Oa Trae00| 0a oes qias0o| —aaoi | = oes Lt jot. Taaoiy esos] | § 88t jp | Faso, oad] |B CT = t Traeoo| onal] FAT t 114700) a Tag “zosoovasro—ravego=—vavton—«121300-—«124600 144900] —- 718000] i | TisT00} | I T 118200) 715300} a= 7844 man 115400) {| 7716500] 0.403 14a 1 lave —Lommeaea]] ¢ PE tz| 115600) 0.407) g 16s. 1 = Tis700| 0.879 a ‘tee00] osi7|_| F145 a) tema tT antes (oo ime tam woo ooo |__| 720200| 0.84 Nour 5 770307 0.844] | 720400] __0.845| [720500] 0.842] 120600] 0.8421 ¥20700{ 0.843] ~ 720500} 0.841 720900] __0.836| yiog| 0.649 721100) 0.843 121200) 0.844 Page 4 JAMIANTIT ATTACHMENT # 2 AFIL Design Department A MASUR, ix Chapter 4 Hydraulics 4.1 HYDRAULIC CHARACTERISTICS ‘The hydraulic characteristics of fiberglass pipe include: * The smooth interior results in low fluid resistence, which could lower horsepower requirements for pumped systems. This characteristic could contribute to a substantial cost savings over the life of a typical piping system. ‘+ The interior pipe surface typically remains smooth over time in most fluid services. Therefore, fluid resistance doos not increase with ag ‘+ ‘The smooth interior allows the pipe diameter to be reduced wi the desired flow. * Inside diameter (ID) is typically larger than IDs of many other pipe materials of the same nominal size and provide a larger pathway for fluid flow. This chapter provides a basis for analysis of the flow capacity, economics, and fluid transient characteristics of fibexglass pipe. 4.2: PRELIMINARY PIPE SIZING ‘The first step in designing » piping system is to determine the pipe size needed to ‘transport a specific amount of fluid. Many engineers have adopted rules that are independent of pipe length but rely on typical or limiting fluid velocities ar allowable pressure drop per 100 ft (80 m) of pipe. After the fluid velocity or the pressure. dvop is mown, it is easy to size a pump to provide the proper flow rate at the required pressure, The following equations are guidelines for the initial sizing of pipe, a 22 FIBERGLASS PIPE DESIGN 4.2.1 Maximum Velocity for Water v= apes (4) Where v= fluid velocity, fs P = fluid density, in = 624 Thi? for water 4.2.2 Maximum Velocity for Corrosive or Erosive Fluids. 24/p 088 (4-2) 4.2.3 Minimum Pipe Diameter for Water d = 0.731QMSGNp >? v8) Where: d = internal pipe diameter, in, @ = flow rate, gpm SG = fluid specific gravity, dimensionless (1 for water) 4.2.4 Minimum Pipe Diameter for Corrosive or Erosive Fluids d= 1.03 (QNSGNp? 4) 4.3 TYPICAL PIPE DIAMETERS ee The equations in Sec. 4.2 represent the minimum pipe diameters or maximum fluid ‘velocities for water and corrosive (or erosive) liquid flow. Typical diameters for fiberglass pressure pipe and suction pipe can be calculated using the fellowing equations. 4.3.1 Typical Diameters for Pressure Pipe Service d = 0821 (QKSGYI G5) 4.3.2 Typical Diameters for Suction Pipe Service = 0.436 (@MSG)I"* 46) 4.3.3 Conversion of Flow Rate to Fluid Velocity v = 0.409 (Qrd*) en “& TpRautics 29 4.4 PRESSURE LOSS CALCULATIONS Head toss, or pressure drop, occurs in all piping systems beoause of elevation changes, ‘urbulence caused by abrupt changes of direction, and friction within the pipe and fittings, number of different computational methods can be nsed to determine the head Joss in fiberglass pipe. The most common methods are the Hazen-Williams, Manning, and the Darey-Weisbach equations. The suitability of each method depends on the type of flow (gravity or pumped) and the level of accuracy required, The relatively, smooth interior surfece of fiberglass pipe should be eonsidered when selecting the roughness coefficient or friction factor in these methods. 4.41 Hazen-Williams Equation ‘The Hazen-Williams equation is applicable to water pipes under conditions of full turbulent flow. Although not as technically correct for all velocities as other methods, the Hazen-Williams equation has gained wide acceptance in the water and wastewater industries because of ils simplicity. Pressure ase cuves fr water based on spect rail of. and viscsty ol t He Pressure Los (P).ps9 per 100 ol pipe r wave 9 8888 8 8 BB! 1,000 2,000 2.000] 10,000 Flow Volume (6), Gatons per Minute Reprinted with permission from Fiberglass Pipe Handbook, Fiberglass Pipe Insttiute, New York, NY. Figure 4-1 Friction loss characteristics of water flow through fiberglass pipe 4 FIBERGLASS P1PB DESIGN ‘The Hazen-Williams equation is frequently presented in nomograph form as Shown in Figure 4-1. Note, however, that graphical solutions usually are valid feo Maier only. When Quids other than water are encountered, a more universal solution Such as the Darey-Weisbach equation should be used. The Hazen-Williams equation nt al for turbulent flow and will ususlly provide a conservative solution Tas determining the head loss in fiberglass pipe. hy = 0.2083 (10010)! (Q#8/at4) (4-8) Where: hy = friction, & Ha0/ 100 ft © = Hazen-Willigms roughness coefficient = 150 (typical value for fiberglass pipe) d = inside diameter (ID), in, NOTE: Graphs and examples use nominal pipe size for simplicity. Tho actual ID should be used in hydraulic calculations, 44.2 Simplified Hazen-Williams ‘Many engineers prefer a simplified version of the Hazen-Williams equation: fie = (42.7 QUC) depts 49) 44.3 Head Loss Converted to Pressure Drop ‘Head loss for any liquid is converted into pressure drop using the fallowing equation: P= (Hp (SO)/ 231 (410) Where: P = pressure drop, psi Hy = hpLI100 SO = fluid specific gravity, dimensionless L = line length, ft New fiberglass pipe has a Hazen-Williams roughness coefficient C value of 150-165. A design value of 150 is frequently used with fibergloss pipe. ‘These values compare favorably to many other pipe materials that have a lower initial value than fiberglass and degrade.over time due ta internal earrusion and scale build-up. The amount of internal corrosion or scale formation is a function of water quality and varies with location and water soures. Example 4-1: Use of the Hazen-Williams equation. Compute the frictional Dressure loss in a 1,500- long, 10-n, diameter fiberglass pipe transporting 2,000 gpm of water, Step 1. Compute the head loss using Rg 49: ‘ay = (42-7) (2,000) (150) (10%) Ay = 1.70 H20/ 100% HYDRAULICS 25 ‘Thon: ‘Total head loss Hy = (1.70) (1,500/100) = 25.6 8 Step 2. Convert head lass to pressure drop using Bq 4-10: P= (25.5)(1.0)/281 = 11 psi 4.4.4 Manning Equation ‘Tho Manning equation typically solves gravity flow problems where the pipe is only partially full and is under the influence of an elevation head only, Qm = (1,486/n) (5) (A) (RPT @1) Where: Qm = flow rate, As S = hydrauli slope fut = (Hi ~H) IE Hy = wpstream elevation, jownstream elevation, ft length of pipe section, f ‘Manning roughness coefficient .009 for typical fiberglass pipe hydraulic radius (A/Wp), cross-sectional area of pipe, ft? Wp = wetted perimeter of pipe, f. oT 4.4.5. Darcy-Weisbach Equation ‘The Darey-Weisbach equation states that pressure drop is proportional to the square of the velocity and the length of the pipe. It is inversely proportional to the diemeter of the pipe. The primary advantage of this equation is that it is valid for all fuide in both laminar and turbulent flow. The disadvantage is that the Darcy-Weishach friction factor is a variable. Once preliminary sizing of the pipe diameter has been completed, the next step is to determine whether the flow pattern within the pipe is Iaminar or turbulent. This characterization of the flow is necessary in the sclection of the appropriate friction factor to be used with the Darey-Weisbach equation, The well- known Reynolds number equation is used to characteride the fluid flow: Re = (D) Wn (412) Where: Re = Reynolds number, dimensionless = fluid kinematic viscosity, 4/5 ID = inside diameter, fe 26 FIBERGLASS PIPE DESIGN TF reennou enna yin Mage 5 ameter inn) =r Peso age zea 44 Bsc uous a ae rereg Ie A se 22) = #2 (vin tees, 0 Reynolds Number FL Reprinted with permission from L. F. Moody, Frietion Factors for Pipe Flow, ASME, 245 B. 47th Sty New York, NY 10017, Figure 4-2 Moody diagram for determination of friction factor for turbutent flow HYDRAULICS 27 ‘This guideline determines the type of flow from the Reynolds mmber: Flow Type Reynolds Nomber Teminar flow Be $2,000 ‘Transition flow zone 2,000 < Re < 4,000 ‘Turbulent flow - Re 24,000 ‘Simply stated, the Darcy-Weisbach equation is as follows: Hy = FL w9/2 ID) g (413) Af flow in the pipe is laminar (i.e., Re S 2,000), the friction factor fi reduces to f= 64/R. (414) Nor Frietion factor for laminar low is denoted as fi and fi denotes friction factor for turbulent flow. : When the flow regime is turbulent (1.e., Re 2 4,000), the friction factor can be determined from the Moody diagram found in most duid mechanics texts (see Figure 4-2), Fiberglass pipe has a surface roughness parameter e equal to 1.7 x 10° fe. When divided by the pipe diameter (e/1D), the friction factor f; for turbulent flow can be extracted from the smooth pipe segment of the diagram. ‘The friction factor for turbulent flow can also be calculated from the Colebrook equation: 1/fP®= ~2 log ((e/ID) 13.71 + 2.51/(Re) i?) (415) Where: fi = Moody friction factor ¢ = surflce roughness parameter 1.7% 10 for fiberglass pipe ‘This equation is difficult to solve because itis implicit inf and requires @ trial and error iterative solution. The following simplified equation relates the frietion factor to the Reynolds nuinber and is accurate to within 1 percent of the Colebrook equation: fe = (1.8 log R/T)? (16) 4.5 HEAD LOSS IN FITTINGS ead loss in fittings is frequently expressed as the equivalent length of pipe that is ‘added to the’ straight run of pipe, This approach has sufficient accuracy for many. applications and is used most often with the Hazen-Williams or Manning equations, ‘The approach does not consider turbulence and subsequent losses created by different fluid velocities. When tabular data aro not available, or when additional accuracy is necessary, head loss in fittings (or valves) can be determined using loss coefficients (K. factors) for each type of fitting. Table 4-1 provides the typical K factors, In this approach the K factor for each fitting is multiplied by the velocity head of the uid flow. Eq 4-17 illustrates the loss coefficient approach, JAMIANTIT ATTACHMENT # 3 AFIL Design Department 4S-oCT-92 TUE 12:25 sHce-supAIL Amiantit Fiberglass Ind. Ltd. P.0. Box 589 Damam 31421 Saudi Arabia ATIN: Mr. Toufic A-Abahussain General Manager suBT: Main Header Inspection Desal Units 21 to 40 sWoc dubai this is to certify that on 18/1/1992, our Maintenance Staff, along vith AFIL's Tnspection Tos, have inspected seawater pipes in Jubail Desalination Plant for Unit 21 to 40 consisting of Fiberglass Reinforced “\ Plastic Pipes, manufactured by amiantit Fiberglass Ind. itd., diameter 1000 m to 2550 mn for weawater intake operating since 1981. at pressure of 8 kg/om2 and ambient seawater temperature, Results of inspection shows no sign of deterioration or damage of any kind. Regards, CadWolMrecd Rolal| acyl hall JA Haba peepee eat ctl Pd pee etl yt yt Boge) gel ritry pla OM ee Lyre de See Seb gs hed Lae 7 wet) WO AL cesta dettl on £ 2 OP AMAKAVA guste, Len ast ae Sib ee Sb et pe WH IY Le i CAPIL 8) pect Lidns Chie eh pa ele eed iT GALS lg ETN ge olen Qe! gles Se (ant) a at peck Settled apo thal ge cate Gia segue geal gS Ughuy LLLaty pYoee = Vere ball ple Se at ag AG let mee Bees Gt pet A nga kay GMA SH gL. pa) olige Bgl ae teu? By) gt ard ot Ls gs Lely gl oe LAs ot ow pe Jus Lae ot SALEH G. AL+ZAHRANY cee Plant Manager siec-Jubail tee Lalas (Sy anmts ( : yale pate (LESAN yet nee, (_ S200/970 pul Lizingi993 Sling Wnter Conversion Corp, “Tels 3610355 Thee 31186 Fux: 3611757 3611689 Tubail - P.O. Box O84 Code 3981 PUVA ET NYY SES API AA KBE ON ree HE TUN) gal 5B Reel gruel Ce. JANMIANTIT ATTACHMENT # 4 AFIL Design Department JAMIANTIT Page 2 For FRP pipes based on experimental study at building research laboratory of Ohio State University, Columbus, Ohio, USA, the absolute roughness (K or <) is found to be 0.00040 ft. i.e. 0.122 mm, where as FRP hand book (5) gives.K as ‘surface roughness parameter to be 1.7 x 10-5 feet for FRP- pipe. This Cloe brook and White equation is in implicit form and therefore require an iterative solution, Wood [2] developed an explicit form of the Cole brook and White’ equation which closely approximates ‘f for Re > 1084 and 105

    You might also like