the challenge of air valves: a selective critical...
TRANSCRIPT
The Challenge of Air Valves: A Selective CriticalLiterature Review
Leila Ramezani, S.M.ASCE1; Bryan Karney, M.ASCE2; and Ahmad Malekpour, A.M.ASCE3
Abstract: One key alternative for removing, preventing, and effectively coping with the often vexing presence of air in water pipelines is thecombination air-vacuum valve. Despite their often effective role, such valves can be highly problematic if not well designed and maintained.This paper critically reviews the current design, application, functionality, and simulation of air valves and the associated shortcomings, with aprimary focus on air/vacuum valves (AVVs). It is argued that the efficient number of air valves along undulating pipelines is yet to be fullyarticulated. Air-valve simulations should expressively consider their dynamic behavior, the physical behavior of any air pockets that mightform below an air valve, and the varying character of the water surface at the air valve location. There is a pressing need for a comprehensiveand systematic study on the proper sizing and location of AVVs for the transient protection of systems. Overall, the efficient application ofAVVs requires broad research and development theoretically (i.e., their physical behavior and improved numerical simulations) and exper-imentally, as well as field studies to document their in situ dynamic behavior and operational efficiency. DOI: 10.1061/(ASCE)WR.1943-5452.0000530. © 2015 American Society of Civil Engineers.
Author keywords: Air management; Air pocket; Air valve; Head loss; Transient; Water pipe; Review.
Introduction
The control of air in water pipes is performed to decrease or preventthe potentially devastating consequences of both mobile and en-trapped air. Among air management strategies, the most popular isthe application of air release/vacuum valves. Air-release valves(ARVs) remove any small-scale air that might occur, while air/vacuum valves (AVVs) generally allow for large-scale removal/admission of air during filling/draining scenarios. Moreover, AVVsare often activated during transient conditions, such as pump fail-ure, admitting air into pipes to counter subatmospheric conditions,and removing the admitted air once pressure increases. Combina-tion air valves (CAVs) integrate the functions of both ARVs andAVVs. However, air valves inevitably play a complex role andcould expose the system to extreme transients and catastrophic fail-ures, particularly if they fail, cease to function properly, or arepoorly sized and located.
Despite their widespread application, quantitative data regard-ing air valve efficiency and effectiveness are scant, and too littleattention has been paid to maintenance and operational issues.Clearly, the proper use of air valves is possible only if such con-cerns are understood and appropriate measures taken to provideefficient design criteria intended to eliminate, or at least reduce,the severity of failures and misbehaviors. This paper argues that
improvement in the design, operation, and maintenance of airvalves can measurably enhance the performance of real-worldwater-conveyance systems.
Current Design Practices of Air Valves
The design or selection of AVVs starts with decisions about theirsize and positioning along the pipeline. Clearly, too few valves willinadequately protect the system and too many represents, at thevery least, a waste of capital and operating resources. However,the choice of appropriate size and location of an air valve is difficultand problematic. Engineers often refer directly to manufacturers orto practical recommendations provided in American Water WorksAssociation (AWWA) M51 guideline for selecting the required sizeof air valves and for deciding about their location along the pipes, aguideline that seeks to provide a basic understanding of the use andapplication of air valves (AWWA 2001). Also, manufacturers’ cat-alogs provide some advice, but it is neither always straightforwardto understand nor clearly justified. In fact, AWWAM51 recommen-dations are those most often used to select the size, location, andtype of air valves, but even these have problematic features.
Sizing Air Release Valves
The first design parameter of an ARV is orifice size. According tothe current AWWA design guideline, the airflow rate, which deter-mines the size of the ARV, is based on the 2% dissolved air in waterat standard pressure and temperature (AWWA 2001). This accountsfor an airflow rate of 2% of the fluid discharge in the system, as-suming that each ARV is to release all the air dissolved in the fluidat any one of the air valve stations. However, the actual dissolvedair in water varies with pressure and temperature, and this potentialamount is almost never released. That is, no particular section ofwater pipes releases this entire amount of dissolved air owing tothe small pressure changes along the pipeline. This implies thatif the first ARValong a pipe releases this amount of air, no air willbe left for release by the second ARV. Therefore, this criterion often
1Ph.D. Candidate, Civil Engineering Dept., Univ. of Toronto, 35 St.George St., Toronto, ON, Canada M5S 1A4 (corresponding author).E-mail: [email protected]
2Professor, Civil Engineering Dept., Univ. of Toronto, 35 St. George St.,Toronto, ON, Canada M5S 1A4. E-mail: [email protected]
3Ph.D. Candidate, Civil Engineering Dept., Univ. of Toronto, 35 St.George St., Toronto, ON, Canada M5S 1A4. E-mail: [email protected]
Note. This manuscript was submitted on August 15, 2014; approved onJanuary 30, 2015; published online on March 18, 2015. Discussion periodopen until August 18, 2015; separate discussions must be submitted forindividual papers. This paper is part of the Journal of Water ResourcesPlanning and Management, © ASCE, ISSN 0733-9496/04015017(11)/$25.00.
© ASCE 04015017-1 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
overestimates the size and number of air valves along a pressurizedmain where pressure is changing. McPherson (2009) andMcPhersonand Haeckler (2012) proposed considering the reduction indissolved-air volume under operating pressures in order to designsmaller air valves. However, the cost effectiveness of this approachrequires a broader study. On the other hand, air in the gaseous phasemay enter the pipe from other sources (e.g., pumps and valves).Hence, this uncertainty and variability in the amount of airmake the design criterion for sizes of ARVs questionable. Consid-ering the uncertainty in the amount of air intrusion into the system,the following questions arise: What airflow rates should be used asdesign criteria for sizing ARVs? Or, more clearly, what size of ARVefficiently removes small-scale air in pipes and enhances systemperformance?
Sizing Air Vacuum Valves for Filling/Draining
Sizing AVVs is a crucial consideration since, for a specified airflow,valve size controls the pressure differential at which the air is ex-hausted. A minimum valve size is established by finding the sizefor filling, which is usually less than the size for drainage. In theinitial filling of pipelines, the airflow rate through the AVVs shouldbe the same as the pipe filling rate (i.e., 0.3 m=s). For a drainingscenario, the rate at which air is admitted into the system should bethe same as that at which the pipeline is drained (i.e., 0.3–0.6 m=s).In practice, pressure differentials of 13.79 and 34.47 kPa (2 and5 psi) are considered for filling and draining, respectively (AWWA2001). The 13.79 kPa (2 psi) differential is a threshold pressure thatmaintains a low air speed in the valve orifice to prevent turbulence,valve slam, and substantial pressure rises, which might lead to pre-mature closure. The 34.47 kPa (5 psi) differential pressure for airinflow is considered a safe pressure threshold to protect pipelinesand gasketed joints from damage due to the effect of vacuumpressures.
Valves are rated according to the maximum pressure underwhich they can operate as well as by orifice size (Thomas 2003).Manufacturers provide their customers with sizing tables or perfor-mance curves with recommended operating pressures. Fig. 1 is aperformance graph, generally provided by a typical air valve manu-facturer, for sizing AVVs based on the filling scenario. Evidently,the orifice size of air valves is designed according to the airflowrate in the system and the pressure differential across the valve.
Generally, in water distribution systems, the recommended outletdiameter of AVVs for a filling scenario is 50–200 mm, while diam-eters of 1–5 mm are suggested for ARVs to release air under pres-sure during normal operating conditions (Stephenson 1997).
Furthermore, ARVs should be sized such that they prevent de-structive transient pressures. Considering the airflow rate undersonic conditions, Bianchi et al. (2007) studied experimentallyand theoretically the appropriate size of air valves during a fillingscenario that would avoid any transient pressures. The basic con-siderations for deriving these practical equations are the maximumallowable overpressure and maximum desired filling velocity. It hasbeen proposed that the most suitable size for air valves under afilling scenario is one in which air is released in a shorter timewhile generating allowable overpressures. For anyone who hasnot directly experienced it, it should be noted that the acoustic im-pact of a valve approaching sonic conditions can be excruciating.Hence, sound should be a consideration in air valve design.
Sizing Air Vacuum Valves for Transient Conditions
As illustrated in the previous section, AVVs are largely sized byconsidering filling/draining scenarios, both of which are largelycontrollable events. Obviously, filling and draining rates can usu-ally be controlled to prevent any consequence of air entrapped inthe system. However, AVVs can have a more crucial role in pro-tecting the system against transient events resulting in vacuum con-ditions. Such a positive role of AVVs is obtained from publishedexperimental data confirming that if air is drawn into a pipe duringnegative pressures, transient surges will be considerably damped(Collins et al. 2012). Vacuum conditions occur in pipes as a resultof sudden transient events, such as power failures at pumps. Therecan be significant consequences to water pipelines in the presenceof such negative pressures. For instance, the resulting negativepressures cause flow disruption and they allow contamination toenter into the pipe. Hence, it is more conservative to size and locateAVVs by consideration of such events. In other words, the sizeof inflow and outflow orifices of AVVs should be selected withcare if vacuum conditions are to be controlled efficiently.
AVVs can protect a system against such negative pressures ifthey are well sized and located. Otherwise, they themselves willinduce secondary transient events, as illustrated later in the paper.For instance, high outflow air valves exacerbate positive transientpressures in systems with negligible amounts of entrained air (Lee1999; Stephenson 1997). The importance of AVV characteristicson transient pressure surges is shown by a case study of a largetwin pump line by Devine and Creasy (1997). A large outletmay result in the rapid expulsion of air and consequent high pres-sures. However, at smaller outlets, the air cavity below the air valvecan cushion the deceleration of the reverse flow. This helps to con-trol the secondary transient pressures (Lee 1999). However, if theoutlet size is too small, air compression occurs below the AVV. Itshould be mentioned that air under pressure contains a hugeamount of energy and can potentially bring about explosive con-ditions. Hence, leaving pressurized air in the system should requirethat the system be designed according to the strict ASME codes forpressure vessels. In other words, discharging air too slowly or try-ing to use a cushioning effect of air will lead to a new set of designcodes, which represents a complication in water pipelines. It is usu-ally advised to properly size AVVs to prevent such explosive con-ditions, and it is concluded that designing water pipelines based onair vessel codes is rejected. Generally, AVVs with low outflow andhigh inflow capacities have been effective at suppressing peak pres-sures and valve slamming as well as reducing negative pressurescaused by transient events (Espert et al. 2008; Lee and Leow 1999).
Outflow Orifice Diameter (in)
1 2 3 4 6 8 10 14 20
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
10000010000100010010Air Outflow through the Large Orifice (scfm)
Pre
ssu
re D
iffe
ren
tial
Acr
oss
th
e A
ir V
alve
(kP
a)
Fig. 1. Performance graph for sizing AVV based on filling scenariogenerally provided by a typical air valve manufacturer
© ASCE 04015017-2 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
Properly sized and located AVVs can potentially serve as acomplementary surge control devices, especially in highly undulat-ing pipelines in which large air vessels must be placed. In suchsituations, placing AVVs in the system may dramatically reducethe size of a large and expensive surge vessel. The effect of AVVson reducing the size of the main system surge control devices issometimes remarkable, as shown in Fig. 2. In this example, anair chamber, located near the pump at the upstream end, is designedfor sudden power failure in both the presence and absence of AVVsat the high point. As shown, the air chamber size, without consid-ering the AVV at the high point, is 200 m3, while the introductionof an AVVat the high point reduces the air chamber size to 25 m3.That is, the presence of an AVVat the high point reduces the size ofthe air chamber by 87.5%. This confirms that AVVs can contributegreatly to reducing the cost of main surge protection devices in un-dulating pipelines.
Furthermore, there is a crucial area ratio between the air valveinlet and outlet if these valves are to play the role of a water hammerprotection device (Zhu et al. 2006; Yang and Shi 2005). Of course,the magnitude of severe pressure peaks following rapid air valveclosure also depends on the location of the AVV, the pipe configu-ration, and the flow characteristics. For instance, when a transientevent occurs upstream of the AVV, the severity of the secondarytransient pressure at the AVV location (Fig. 3) depends on the in-tercepted wave height (i.e., HP) at the high point. In this figure,ΔH1 is the reduced transient pressure wave induced upstreamof the AVV. QP=Q0 is the ratio of the reduced discharge afterthe wave arrives at the high point. As shown, for higher elevationsof the high point, the intercepted wave height is reduced, QP=Q0 isincreased, and therefore the secondary transient pressure is higher
0
10
20
30
40
50
60
70
80
90
100
110
120
0 1000 2000 3000 4000 5000 6000Distance (m)
Pre
ssu
re H
ead
(m
)
Hmax with AVV, air chamber size=25 m3 Hmin with AVV, air chamber size=25 m3 Pipe Elev. (m)Hmax without AVV,air chamber size=200 m3 Hmin without AVV,air chamber size=200 m3 Hmax without AVV,air chamber size=25 m3 Hmin without AVV,air chamber size=25m3
AVV
Air Chamber
Fig. 2. Effect of presence of AVV at high point on reducing air chamber size (87.5% reduction in chamber size)
HP
V=0
a -ΔH1
AVV
(a)
(b)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Hp/ΔH1
ΔHm
ax/ Δ
H1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
QP/Q
0
DHmax/DH1QP/Q0
Fig. 3. Effect of intercepted wave height on secondary transientpressures at AVV location (data from Ramezani and Karney 2013):(a) system layout, sudden pressure drop at upstream valve; (b) effectof intercepted wave height at high point (HP=ΔH1) on secondarytransient pressures (ΔHmax=ΔH1) at AVV location
© ASCE 04015017-3 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
(Ramezani and Karney 2013). However, a systematic study of theeffect of the size and location of AVVs on secondary transient pres-sures has yet to be published. That is, current studies are performedwithin a single system configuration, but they have not explored thelocations where the proper sizing of AVVs are most or least crucialin transient events.
Location of Air Valves
The second design parameter of air valves (i.e., ARVs, AVVs, orCAVs) is their location and distribution along the pipeline. AWWArecommends installing air valves at intervals along ascending,descending, and horizontal lines (i.e., intervals of 1,500–2,500 ft)and changes in grade as well as all the high points along the pipe.However, this is at times a conservative approach, and it is seldomspecified how critical each location along a pipe profile is. Elevatedhigh points are naturally the most common places that ARVs orAVVs are recommended in both North American (AWWA 2001)and European guidelines (DVGW 2005). Characterizing criticalhigh points to install ARVs and AVVs or CAVs requires knowledgeabout the potential of air accumulation at the high point and thetransient behavior of the system, respectively. However, there islittle published work regarding the influential parameters on air ac-cumulation along pipelines or at high points in pressurized systems.Moreover, little work has been done regarding the locations alongpipes where appropriate selection of AVVs is crucial for alleviatingtransient induced pressures.
Placing ARVs at improper locations can be problematic ratherthan helpful. Among the improper locations of ARV placementare the high points running under siphon conditions. Obviously,the existence of a permanent vacuum condition at a high point(e.g., with a siphon) may restrict the application of ARVs at thoselocations. According to catalogs of air valves, ARVs can potentiallyoperate under vacuum conditions admitting air into the system.Hence, placing such valves at siphons where permanent vacuumconditions are actually desirable may lead to flow disruption owingto breakages of siphon flow. In such cases, the location of the highpoint relative to the downstream boundary affects the magnitude ofthe discharge reduction.
The consequence of placing ARVs at siphon locations is easilyillustrated. Fig. 4 shows a configuration in which placement of anARVat the high point, previously running under vacuum conditions(i.e., a siphon), breaks the siphon flow. Obviously, siphon flow canbe calculated by writing the Bernoulli equation between Points 1and 2 as well as 1 and S (Fig. 4), resulting in Eqs. (1) and (2),respectively. However, at the high point, the existence of anARV that could activate during vacuum conditions may introduceair into the system, increasing the negative pressure in pipeline tothe atmospheric pressure and reducing the design discharge. Here,the governing equation is Eq. (2). The reduced discharge can becalculated by considering the atmospheric pressure at the highpoint and solving for flow velocity using Eq. (2). The pipe diameter(m), pipe length (m), design discharge (m3=s), friction factor, andK1 are 0.5, 2,000, 0.328, 0.017, 0.5, and 1, respectively. The
change in the relative design discharge (Q=Qd) with the locationof the high point is shown in Fig. 5, in which Q is the reduceddischarge and Qd is the design discharge. As shown, the flow re-duction is affected by the location of the high point relative to thedownstream boundary. The closer the high point is to the down-stream boundary, the more flow reduction occurs by admittingair into the system. Such effects are more pronounced in shorterpipelines. To illustrate, Fig. 6 is obtained by considering a constantpipe length upstream of the high point in Fig. 4 and altering thedownstream pipe length. As seen in Fig. 6, as the downstream pipelength decreases, the discharge has lower values, implying morereduction in design discharge
Z1 − Z2 ¼V22
2gþ f
LDV22
2gþ K
V22
2gð1Þ
Z1 − ZS ¼PS
γþ V2
2
2gþ f
L1−SD
V22
2gþ K1
V22
2gð2Þ
To prevent such a reduction in the design flow at siphons,AWWA (2001) recommends placing a vacuum check device atthe outlet of the ARV. Although manufacturers claim to providevacuum check valves, their function and practical efficiency insiphon conditions remains questionable. Even if these devicesare effective at preventing siphon breaks, any malfunction of suchdevices could result in a reduction in the design discharge and thehydraulic efficiency of the pipeline. Hence, placing ARVs at suchlocations requires a detailed study of such potential effects and con-sideration of the associated risks.
1
2S
L1-S LS-2
Qd
Fig. 4. Breaking siphon condition as a result of admitting air intosystem by ARV at high point
0.70
0.75
0.80
0.85
0.90
0.95
1.00
0.2 0.3 0.4 0.5 0.6 LS-2 /L1-2
Q/Q
d
Fig. 5. Reduction in design flow from breaking siphon condition byARV at high point (Q = reduced discharge; Qd = design discharge)
0.6
0.7
0.8
0.9
1.0
0.2 0.5 0.8 1.1 1.4 1.7
LS-2 /L1-S
Q/Q
d
Fig. 6. Ratio of reduced discharge to design flow; design dischargereduces owing to breaking of siphon condition by ARV at high pointas in Fig. 4 (Q = reduced discharge; Qd = design discharge)
© ASCE 04015017-4 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
Type of Air Valve
The third design parameter is the type of air valve. After selectingthe appropriate size and location of air valves, the designer shouldselect among a variety of types of air valve supplied to the marketby several manufacturers. While applying the same hydraulic andmechanical principles, these valves are different in physical proper-ties (i.e., material, shape, float arrangement) and performancecriteria with optional add-on features such as surge relief or anti-slam devices (McPherson 2009). For example, while AWWArecommends air valves made of cast iron, all manufacturers offerdifferent material bodies, such as ductile iron, stainless steel, caststeel, and cast bronze, as well as polypropylene and reinforced ny-lon. Obviously, the capital cost of these valves differs but the main-tenance cost is unknown and the relative merits are often unclear(Radulj 2007). It is seldom obvious to designers what characteristicparameter of these valves (e.g., working pressures, antislam fea-tures, or fewer reported failures) gives them priority over othersin a specific water main. In other words, the influence of suchcharacteristics on system performance is not yet fully reported.Furthermore, there is a gap in the literature regarding whether thesedifferent types of air valve may affect the overall and proper num-ber and size of air valves in a given system. Overall, the priority ofone type of air valve over others is still contested.
Operating Problems of Air/Vacuum Valves duringAir Release
The main function of AVVs is to expel excess air from the systemwhen it is experiencing high pressure or to admit air into the pipeunder subatmospheric conditions (Wylie and Streeter 1993). How-ever, analysis of the transient behavior of systems protected byAVVs by developing and applying either numerical models (Leeand Leow 1999) or simplified water hammer equations (Lingireddyet al. 2004; Li et al. 2009) reveals that rapid expulsion of air throughair valves causes severe transient pressures referred to as secondarytransient pressures.
Secondary Transient Pressure during Air Release
A secondary transient event is caused by rapid air expulsion andsubsequent rapid valve closure. When water strikes a closed airvalve, it moves at the same rate as the exhaust air, leading to severeflow deceleration and high pressure surges inside the valve. This ismost common both during filling scenarios and after suppression ofvacuum conditions while the incoming air from the AVVs is beingreleased from the pipe. Secondary transient pressures have beenstudied experimentally during filling scenarios in a simple horizon-tal (Zhou et al. 2002a, b; Martin and Lee 2012) or a more complex(De Martino et al. 2008) piping system. Peak pressures up to 2.7and 15 times the upstream absolute head were observed at values ofd=D < 0.086 and d=D ¼ 0.2 (where d is the air valve diameter andD is the pipe diameter), respectively (De Martino et al. 2008; Zhouet al. 2002a, b). In addition, Martin and Lee (2012) observed thatmaximum pressures at the air valve orifice occur at d=D ¼ 0.18and are affected by both the orifice size and the ratio of the absoluteupstream reservoir pressure to the initial absolute air pressurenear the valve in a simple horizontal reservoir-pipe system. Theyconcluded that initial air volume at the valve is less influential onpeak pressure compared with the two aforementioned factors. Also,numerical results of secondary transients have been presentedduring rapid shutdown and subsequent start-up of a pumping linefor an undulating pipe containing an air valve at the high point(Lingireddy et al. 2004). A simple predictive formula is also
available for calculating the maximum pressure based on Allievi-Joukowski results (De Martino et al. 2008). Secondary transientscan lead to air valve breakage. For instance, such frequent highpressures have caused breakage in the inlet or outlet sections ofair valves in Pinellas County’s wastewater force main (Li et al.2009).
Other conditions for creating secondary transients are met whenthese valves do not operate well and premature closure occurs(Lee and Leow 1999; Lingireddy et al. 2004; Li et al. 2009). Forinstance, conventional air valves can close prematurely at surpris-ingly slight pressure changes (Thomas 2003), and it is difficult tolimit pipeline pressures to below-closure thresholds.
Pressure Oscillation Patterns during Air ReleaseDepending on the size of air release orifice, three types of pressureoscillation patterns occur in entrapped air pockets during the rapidfilling of a pipeline (Zhou et al. 2002a), and there are two phasesin each oscillation pattern: a low-frequency pressure oscillation(i.e., rigid column behavior) followed by a sudden pressure peak(i.e., water hammer behavior) when the water column first reachesthe orifice (De Martino et al. 2008; Zhou et al. 2002a). The durationof the first phase is reduced with an increase in the driving head andorifice size and increases with air pocket size (De Martinoet al. 2008).
For small orifice sizes, an air cushioning effect is dominant andthe period of pressure oscillation is long. For large orifice sizes, theair cushioning effect is negligible and the water hammer effect isdominant. In this case, there is no long-period pressure oscillationpattern and just a water hammer impact pressure is present. Forintermediate orifice sizes, the air cushioning effect begins to dimin-ish and the water hammer pressure becomes more pronounced asthe orifice size grows. Therefore, there is both a long-period oscil-lation followed by a short-period oscillation pattern. Experimentsshow that the peak pressure occurs before the final air residual isremoved from the pipe owing to air pocket compression (DeMartino et al. 2008; Martin and Lee 2000).
Influential Parameters on Secondary Transient PressuresAs illustrated by a simplified analytical equation by Lingireddyet al. (2004), the magnitude of secondary pressures depends on pipeand air valve characteristics and the pressure inside the valveimmediately before the final air release. Also, experimental databy Zhou et al. (2002b) and De Martino et al. (2008) show thatthe upstream driving head, air pocket volume, and orifice sizeof air valves are the most influential factors on maximum pressure.To illustrate, the pressure peak increases with the upstream drivinghead and outlet orifice diameter. The maximum pressure occurs at acertain ratio of orifice diameter to pipe diameter (i.e., d=D). Thevalue ascribed to this ratio varies among researchers. Martin andLee (2000) reported a value of d=D ¼ 0.14 and later observed peakpressures at d=D ¼ 0.18 (Marin and Lee 2012), while Zhou et al.(2002b) reported a value of d=D ¼ 0.2 where the maximum pres-sure peak occurred. This difference may be due to different con-ditions under study. For example, different experimental setups,such as pipe length and diameter, water column length, orificediameter, and driving head, are factors that were different in eachstudy (Zhou et al. 2002b; De Martino et al. 2008). After reaching acertain orifice size, the peak pressure decreases as the orifice sizeincreases (Martin and Lee 2000; Zhou et al. 2002a, b, 2004; Martinand Lee 2012).
Overall, since the d=D criteria for the occurrence of peak pres-sure are different in each study, these data cannot be generalized.Furthermore, this phenomenon is given little physical interpretationin the literature. For example, the literature contains no discussionas to why an air pocket’s cushioning effect is less influential at
© ASCE 04015017-5 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
lower driving heads and smaller air release rates, although peakpressure generally decreases with air pocket volume. In addition,almost all of these criteria are based on assuming a rigid water col-umn theory without considering friction and elastic effects. Hence,the effects of water elasticity on such criteria and pressure oscilla-tion patterns are yet to be investigated.
Real-Time Dynamic Behavior of Air/Vacuum Valves
The dynamic behavior of air valves affects the system’s transientresponse during air release from the pipe. To illustrate, the time ofclosing and opening of an air valve, the amount of residual air be-low the air valve before it closes, and the amount of time it takes toclose are defined as the dynamic characteristics of air valves. Suchparameters are difficult to measure in the laboratory and, therefore,involve some uncertainty. For instance, the discharge coefficient ofair valves is determined through standard published tables (AWWA2001), and there is considerable uncertainty in these values. Suchdynamic behavior is different for each type of air valve (i.e., its floatshape, weight, and size affect the closing behavior of the valve).Pipe layout, the severity of transient, and type of air valve instal-lation contribute to this as well.
The dynamic behavior of air valves is rarely systematicallystudied or modified. The main reason for this lack of attentionis a poor understanding of the dynamics of air valves owing to scantlarge-scale experimental and field data. Also, the complexity oftwo-phase flow characterization might explain the scarcity of pub-lications on this topic. Although the effect of such dynamic behav-ior is not comprehensively understood, some experimental workshave been published on this topic. It is noteworthy that scale effectshave not been explored in experimental studies, and therefore, theycannot be generalized for all types of air valves. Yet, these explo-rations have important implications for the dynamic behavior of airvalves.
Recent work attempts to provide technical information and tounderstand the behavior of air valves having different geometrical,mechanical, hydraulic, and dynamic characteristics. Because sizingcharts provided by manufacturers are unreliable, initially the char-acteristic curves of air valves must be tested and compared with thecharacteristics provided by manufacturers (Lucca et al. 2010). Thiscan lead to more realistic simulations of air valve behavior in pipe-lines. For example, Bergant et al. (2012) determined the dynamicperformance of float-type air valves (e.g., metal ball or plastic cyl-inder floats) by measuring their response to flow acceleration anddeceleration in a large-scale pipeline for three scenarios of pumpstart-up and shutdown and pipe rupture. They presented the rela-tionship between flow rate and pressure drop across an air valve forseveral float positions.
Air Release or Admission through Air Valves ConsideringDynamic BehaviorGenerally, air release through AVVs is divided into two stages(Bergant et al. 2012). In the first stage, water and air acceleratetoward the air valve until almost all the air is removed from the pipe(i.e., initial event). In the second stage, residual air compresses untilhigh pressure occurs and the water column stops and reverses(i.e., air compression event). The flow velocity at the end of theinitial event is referred to as the terminal velocity, which is themaximum value of the water velocity. However, according toreal-time experiments on different models of air valves, some airvalves close before air is completely exhausted from the pipe(Bergant et al. 2012), while others close only after a certain amountof water is discharged. The volume of residual air depends on thedynamic behavior of the air valves (e.g., closing times), the pipelayout, the severity of the transient, and the air–water mixing
process. Also, a longer closing time means less residual air.Furthermore, the higher the impacting water column, the smallerthe residual air (Carlos et al. 2011).
Following valve closure, the residual air continues to compressuntil the water column is arrested and then reverses. A maximumpressure spike occurs at the time of flow reversal (Bergant et al.2012). The residual air is an influential parameter on transient pres-sures, especially at start-up events. Therefore, the most dangeroussituations are those in which small uncontrolled volumes of air re-main inside the pipe and cannot be removed properly through an airvalve (Bergant et al. 2012; Carlos et al. 2011; Arregui et al. 2003).Results show that the smaller the residual air below the air valves,the higher the maximum pressure. Moreover, experimental datashow that pressure peaks are inversely proportional to the closuretime of air valves up to a certain limit, after which peak pressure isnot dependent on closure time (Arregui et al. 2003). However,Carlos et al. (2011) showed that in real systems, the closing timeof air valves is not always a significant factor for the pressure peakmagnitude. Additionally, peak pressures are inversely proportionalto moving water column inertia and directly proportional to termi-nal velocity (i.e., velocity of water at the time of air valve closure)(Arregui et al. 2003).
Overall, the effectiveness of air valves for surge protection de-pends on several factors, such as system configuration, the physicalproperties of the pipeline and fluid, the characteristics of the airvalves, and the air distribution in the system (Lee 1999). However,the complexity of the characterization of two-phase flows and airvalves has limited the associated theoretical works. Hence, re-searchers have investigated the dynamic performance of air valvesand their influence on transient pressures by performing large-scaleexperiments (Bergant et al. 2012; Carlos et al. 2011; Arregui et al.2003; Cabrera et al. 2003).
According to such experimental results, air valves do not re-spond as quickly to vacuum conditions as has traditionally beenexpected in the literature (Bergant et al. 2012; Cabrera et al. 2003).That is, air valve opening occurs with a small delay (on the order of20 ms). This makes the occurrence of cavitation inevitable. Butafter a small delay in the admission of air, the pressure surges arestrongly reduced. There is also a time delay between the air valveclosure and pressure rise due to cavity collapse (Bergant et al.2012). Such a delay could occur if the valve body stuck to the valveseat. Yet these valves have proven to be beneficial as surge controldevices if properly sized and located. Experimental results showthat the maximum air pressure at the air valve location dependsstrongly on air valve size, and both larger and smaller sizes of airvalves exacerbate transient pressures. Usually, large sizes of airvalves result in high pressures as a result of the high velocity ofthe adjacent water column. And smaller air valve sizes give riseto peak pressures as a result of excessive compression of air pocketsbelow air valves (Cabrera et al. 2003). Hence, this earlier work con-firms that the choice of an appropriate air valve size should be madeif excessive pressure surges are to be avoided. In addition, exper-imental data imply that the discharge coefficient of air valves playsan important role in peak pressures. According to experiments sim-ulating the expulsion of air through several commercial air valves,the smaller the discharge coefficient, the smaller the impact veloc-ity of water arriving at the air valve at the time of air valve closureand the lower the resulting peak pressure. Peak pressures aregreatly affected by the amount of residual air below the valve.Small air residuals in the pipe can be dangerous and burst the pipe.The friction factor and air valve outlet coefficient are other keyparameters since they affect the impact velocity of water. For in-stance, a smaller air valve outlet coefficient controls the terminal
© ASCE 04015017-6 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
velocity of water when it hits the closed air valve and then reducesthe corresponding peak pressure (Carlos et al. 2011).
Therefore, according to the available experimental and theoreti-cal work, the effectiveness of AVVs in surge protection depends onfactors such as the size and discharge coefficient of the air valves.Water hammer protection of air valves is mostly dependent uponthe optimal value of the outlet to inlet ratio. Certainly this dependson system configuration (Zhu et al. 2006). A case study of a watersupply in Riyadh, Campbell (Lee and Leow 1999) showed thatnonreturn air valves cushioned the impact of slams from negativesurges, allowing a substantial reduction in the size of surge vessels.However, what is not obvious is the degree to which the installationof air valves might protect the system against vacuum conditions ifthey are installed in pipelines as the main surge control devices orsupplementary surge protection measures. Furthermore, the uncer-tainties associated with the design parameters of air valves (e.g., dis-charge coefficient or even their basic functioning) could affect theresults. But such effects have not been considered in the publishedtheoretical works.
Air Valve Chattering due to Secondary Transient PressuresWhile air is released through the air valve, it compresses and accel-erates toward the air valve. Simultaneously, the adjacent water col-umn is accelerating. High pressure peaks can result either from thecompression of air pockets or during air release when the acceler-ating water column impacts the closed air valve. The reduction ofvelocity in a water column can produce secondary transient pres-sures on the order of 1–10 bar or higher, and reflections of suchpressure waves can potentially cause cavitation in the pipe. Suchtransient events can cause air valves to repeatedly open and closewith high frequency, a phenomenon called air valve chattering(Bergant et al. 2012). The water hammer pressure and the pressuredue to air pocket compression creates high pressure peaks, and airvalve failure can lead to column separation along the pipeline(Bergant et al. 2004). Obviously, the location of air valves relativeto the hydraulic grade line (HGL) and the downstream boundary(e.g., a reservoir) affects the frequency of chattering. That is, forany vertical location of the high point (i.e., air valve location),if the total pipe length downstream of the air valve is short, thenthe wave round trip is small. In these situations, less air enters thepipe through the air valve to prevent vacuum conditions, and lesstime is required to discharge the previously admitted air and tocreate the secondary transient. Hence, chattering occurs with morefrequency at the air valve location. Such issues should be consid-ered in the design stage of air valves.
Overall, operational problems are inevitable, and proper pipe-line venting can take place only if AVVs are sized and located ap-propriately throughout a system and if they resume operationduring transient induced pressure spikes, and these are subjects thatrequire more systematic research. And since water systems are nobetter than their weakest component, a more rational approach tothe design and selection of air valves is urgently needed.
Maintenance Problems
Air valves require at least annual maintenance and inspection fordamage. In one European guideline, PED 97/23/EC (EuropeanCommission 1997), AVVs are considered safety valves, and there-fore their maintenance is emphasized. Regular maintenance andfunction testing are also advised in the German guideline (DVGW2005). However, due to the relative inaccessibility of most AVVlocations in vaults requiring confined-space-access procedures,the valves are often left exposed to weathering, physical damage,and wear, which eventually render them nonoperational. Also,buried valves, located below manholes to improve accessibility,have the potential to become inaccessible due to flooding or siltingof the utility vaults below the manholes (Fig. 7). For example, anassessment of the functionality of air valves existing over 30 yearsin the city of Houston water transmission main revealed that half ofthe air valves were not functioning (the total number of air valveswas 31) and half of the buried air valves were inaccessible or hadleaky gate valves and could not be tested for damage (Gregoryet al. 2006).
The associated cost of flooded chambers [Fig. 7(a)] adds to themaintenance cost of air valves. Also, contaminated water mightenter the pipes through leaking air valves during a negative pressureevent, creating health risks (Besner et al. 2010). Consequently, airvalves add to the maintenance cost of a system if they are not wellused and maintained. Fig. 7 shows evidence of poorly maintainedair valves, which has led to nonoperational valves.
Avoiding Operational and Maintenance Problems
The potential of air valves to induce secondary transient events inpipes has provoked Li et al. (2009) to theoretically investigate amodern configuration of air valves: installing an air-throttling or-ifice at the air valve outlet, an antislam device, and an air accumu-lator that reduce the flow velocity at the valve seat and the resultingsecondary pressures. Air-throttling orifices prevent all the air from
Fig. 7. Flooded and poorly maintained air valves [(a) reprinted with permission from Besner et al. (2010); (b) reprinted with permission fromEscarameia (2005), Copyright © HR Wallingford 2005]
© ASCE 04015017-7 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
being expelled, and therefore, a portion is trapped in the valvechamber. The greater the air accumulation, the more the pressuresare dampened. However, this raises concerns regarding pressurespikes that can be created by such trapped air pockets during sud-den pump start-up. Although this method was tested under two sce-narios of shutdown and the following start-up, the start-up period(t ¼ 300 s) was 60 times longer than the shutdown duration(t ¼ 5 s), leading to a less severe induced transient. Obviously, careshould be taken to consider sudden start-up cases, resulting insevere transients, while designing the air accumulator capacityin the valve. Although the effects were theoretically investigatedthrough a case study, the real-world practical efficiency of thesevalves has not yet been published.
Another approach to minimizing operational problems causedby secondary transient pressures at AVVs is the use of standpipesbelow the AVVs. Here, the main idea is that the pipe diameter islarger than the standpipe, and consequently, a water hammer wavewill be attenuated when it reaches the main pipe. Properly sizedstandpipes can suppress transient pressure caused by air valveslamming. Generally, parameters affecting a water hammer arethe standpipe length and diameter, the ratio of pipe to standpipediameter, and air valve size, as well as the relative residual airvolume with respect to standpipe volume in the standpipe whenthe valve is closed (Stephenson 1997). Thus, the available numeri-cal studies indicate that the sizing of standpipes should be per-formed with care. However, their effectiveness in real systems isstill questionable and has yet to be published.
Additionally, automatic air valves, especially those applied towastewater systems, are uncertain and unreliable since they requireregular maintenance, and the possible malfunction of these valvescan create high pressures that endanger pipelines (Tchobanoglousand Metcalf 1981). However, manufacturers claim that some ofthe features of modern automatic air valves allow them to functionefficiently while requiring simple and effective maintenance(Zloczower 2010). Among such features are conical bodies witha large midriff that encompasses a larger initial air pocket at thevalve body, causing the air pocket to compress at high pressuresand resulting in a constant water level inside the valve, preventingthe sealing mechanism from clogging and leakage. Also, outward-slanting valve walls and a funnel-shaped bottom prevent the accu-mulation of grease by redirecting it to the force main. Furthermore,the previous complex and problematic sealing mechanism with in-ternal levers, hinges, and pins is substituted with a freely movingfloat and simple rolling seal mechanism, which provides a larger airrelease orifice and works even at low pressures [20.68 kPa (3 psi),2 m]. More importantly, a light body, nonslam accessories, and ahydraulically piloted diaphragm instead of a float that can poten-tially eliminate local surges due to air valves are embedded in thesevalves (Zloczower 2010). Of course, the dynamic behavior of thesevalves has not yet been studied, and their practical efficiency hasnot yet been published. Also, in wastewater systems, such valvesmay malfunction as a result of clogging due to floating dirt andgrease (Tukker et al. 2013). Therefore, the idea of little mainte-nance as claimed by manufacturers can be rejected.
Of course, the most common recommendation in the literatureto avoid secondary pressures is to use smaller outflow sizes ofair valves (Lingireddy et al. 2004; Zhou et al. 2002b). Such findingsin the literature emphasize that each specific system should beanalyzed in order to select an appropriate outflow air valve size.However, there is a gap in the literature regarding the effect ofAVV location on its sizing. In other words, rules of thumb for se-lecting an appropriate outflow size of an AVV according to its lo-cation are not yet available. Also, there are few studies concerningeither the priority of each of the discussed strategies over the others
or of the best possible or practical approach to avoid operationalproblems arising from secondary transient pressures at AVVs.
Therefore, AVVs and ARVs, though designed to reduce the dev-astating problems created by vacuum conditions and the presenceof air pockets in pipes, also add to the operating and maintenanceproblems in water systems. To designers, reducing the relatedoperating and maintenance problems justifies eliminating even asingle air valve from the system. In addition, the German guideline(DVGW 2005) encourages restricted use of AVVs in drinkingwater pipelines due to the possible intrusion of contaminationand the problems that the sucked-in air, which is possibly trans-ported downstream of the AVV, can create in pipelines. But howsmall a number is reasonable is still a key design question.Certainly, no attention has been paid to such a strategy in the liter-ature, and the great concern is reduced system performance result-ing from the removal of a specific air valve along a pipe. This leavessystem owners wondering how effective these air valves are at im-proving system performance. If an air valve in a specific location ofthe pipe profile has little effect on system performance, avoiding itwill have the potential to reduce such operation and maintenanceissues.
Basically, reviewing and evaluating the current design criteria ofair valves (e.g., size, location) could play an important role in high-lighting the potential of any possible changes in the design of airvalves along a pipeline. Such analysis would reveal critical loca-tions for installing air valves and allow one to explore how systemperformance changes if an air valve, or a series of air valves, fails tooperate. Consequently, the potential to avoid installing a number ofair valves will be realized. This provides a framework for decisionmakers to determine the reasonable size, type, and location of airvalves according to the desired system performance. Moreover, tobetter assess the performance of air valves and their location andsizing criteria, supervisory control and data acquisition (SCADA)systems should monitor air valve open/close positions and differ-ential pressures (McPherson 2009). Such monitoring may facilitateexploration of the location of the most effective air valves. Further,it would help in identifying those air valves that require regularmaintenance. Consequently, the aforementioned operational andmaintenance problems are reduced through the elimination ofunnecessary air valves. Overall, if air valves are to be neglected(Fig. 7) and if it is proven that they have little effect on systemfunctionality, why not avoid them in the design stage?
Certainly, such an evaluation requires an understanding of theconsequences associated with air intrusion into a system, air accu-mulation at high points, vacuum conditions, and the absence ormalfunction of air valves, as well as a subsequent considerationof system performance, all of which still requires in-depth research.For example, to evaluate current size and location criteria of ARVs,it is essential to know the amount of air accumulation at locationsalong the pipe and the associated consequences (i.e., head loss).According to the available experimental data, head loss createdby air pockets is a function of airflow rate, pipe slope, and watervelocity (Pothof and Clemens 2010; Lubbers 2007; Lubbers andClemens 2005). However, the accumulated air pocket volumeand the relationship between air pocket volume and head lossare not available for pressurized lines. Also, experimental observa-tions have indicated that air accumulation depends on airflow rate,but the associated relationship is not reported in the literature.Therefore, the lack of such data in the literature limits the evalu-ation of design criteria for ARVs. Obviously, the reported resultscould be more effective for such an evaluation if more experimentaldata (e.g., volume of accumulated air at high points, relationshipbetween air pocket volume and head loss) were recorded. Onthe other hand, most of these experimental works were performed
© ASCE 04015017-8 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
under the operating conditions of sewer systems (i.e., low ranges ofpressure: 0–3 bar). Hence, the data could be more reliably appliedto water transmission systems if the experiments were performedunder higher pressurized conditions and if scale effects were pro-vided. Consequently, more research should be done on this subjectif evaluation of design parameters of ARVs is desirable. Also, theevaluation of design criteria for AVVs requires a systematic studyof the transient response of water systems in the presence of AVVs.Of course, the lack of such studies is an important gap in theliterature.
Numerical Modeling of Air Valves: A Critique
Modeling air valve boundary conditions may be handled within theusual framework of the characteristics method. Complete equationsused for air valve boundary simulation are presented in Wylie andStreeter (1993). The mass flow rate of air through a valve dependson the values of atmospheric absolute pressure, p0, and absolutetemperature, T0, outside the pipe, as well as the absolute temper-ature, T, and pressure, p, within the pipe. Four conditions—airinflow and outflow through an air valve at subsonic and soniclevels—are considered. Also, four assumptions are made in the ana-lytical equations for air valve boundary conditions (Wylie andStreeter 1993): (1) isentropic flow is assumed for airflow throughthe air valve; (2) the air mass within the pipe is assumed to obeythe isothermal law; (3) the air admitted into the pipe is assumedto stay near the valve; and (4) the water surface elevation is assumedto remain constant, and the volume of air is considered small com-pared with the liquid volume of pipeline reach.
However, in practice, these four assumptions for air valve boun-dary conditions cannot be universally met. For example, if an airvalve is placed along a sloping pipe, the assumption that the air willremain close to the air valves cannot be satisfied unless a stand pipe isused and the admitted air volume is small enough to stay in the stand-pipe. Otherwise, the pipe should be designed so that it contains highpoints (Zhu et al. 2006). If none of these situations apply, consideringthe air transport mechanism is essential for an accurate numericalsimulation of air valves. Also, the constant water surface at theair valve location is an assumption whose effect on the transientanalysis of the system and the final reverse flow should be examined.
The effect of assuming a constant water surface at an AVV lo-cation on the transient analysis of a system is remarkable. To illus-trate, when a power failure occurs, the AVVat the high point startsto operate, admitting air into the system to prevent column sepa-ration. During deceleration of the forward flow downstream of theAVV, the water level drops at the high point. After forward flowstoppage, reverse flow accelerates toward the AVV and the waterlevel at the high point rises. If the effect of the moving boundary atthe AVV location is ignored, the forward flow deceleration rate andthe reverse flow acceleration rate are underestimated, which meansthat the computed time of cavity growth and collapse at the highpoint will be inaccurate.
Furthermore, in situations where the high point is higher thanthe downstream reservoir head (Fig. 8), neglecting the falling waterlevel at the high point will prevent reverse flow occurrence owing tothe negative reverse driving head. An example is shown in Fig. 8.The inlet and outlet diameters of the AVV (m) and the inlet andoutlet coefficients of the AVV are 0.2, 0.2, 0.6, and 0.6, respec-tively. Consequently, the numerical solution based on such a con-stant water level assumption results in a permanently growingcavity volume at the high point (Fig. 9). Hence, in such conditions,obtaining meaningful numerical simulation results depends on con-sideration of the moving boundary conditions at the AVV location.
Overall, in the literature, less attention has been paid to suchshortcomings in the numerical simulation of air valves. Suchnumerical issues and the previously discussed dynamic behaviorof air valves imply that further research and improvement are stillrequired in the field of air valve numerical simulation. Obviously,real-world data acquisition would be useful in the development ofnumerical models and in the validation of the numerical data.
Summary and Conclusions
In the literature, one air management strategy involves the appli-cation of air control devices such as air valves (i.e., air releasevalves and air/vacuum valves). This paper presents a critical reviewof the use and application of air valves. Areas of improvement inthe literature are discussed and the gaps examined. Of course, themain focus of the paper is on the application of AVVs. Further-more, design guidelines and operating and maintenance issues,as well as the current alternatives for preventing such problems,are discussed. It is concluded that despite the effective role ofair valves in air management, there are very little data about theirfrequency and efficiency of functioning, which is necessary infor-mation for better sizing and positioning and a more efficient appli-cation of air valves.
Although there are several case-specific studies on sizing airvalves based on transient conditions, a comprehensive and system-atic study on air valve sizing has yet to be published. There remain
Elev. 0 m
Elev. 57 m
L=3000 m
L=5000 m
L=3000 m
Elev. 40 mQ0=1 m3/s D=1 m f=0.017
85.19 m
AVV @ Elev. 60 m
Fig. 8. Layout of a pumpline containing a high point higher than thedownstream reservoir head
0
50
100
150
200
250
300
0040020t (s)
H (
m)
Or
Vai
r (m
3 )
0
0.2
0.4
0.6
0.8
1
1.2
Qo
ut (
m3 /s
)
HVairQout
Fig. 9. Variations in head (H) and air volume (Vair) at high point andwater discharge (Qout) at downstream limb of high point in Fig. 8
© ASCE 04015017-9 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
gaps in the literature concerning proper location and, consequently,the efficient number of air valves required along an undulating pipe-line. On the other hand, the available limited experimental studies ondynamic behavior of air valves are presented. Further, the need tomodify the numerical modeling of air valves based on such dynamicbehaviors is highlighted. Obviously, such considerations may influ-ence the proper selection of air valve size, number, and location.
Overall, considering the operational and maintenance issues re-lated to air valves, efficient application of such devices requiresmore broad-based research and development both in a theoreticalcontext (e.g., understanding air valve physical behavior andmodifying air valve numerical simulations) and in experimentalor field studies (e.g., understanding air valve dynamic behaviorand operational efficiency). Obviously, the experimental or field datacan serve as a good support for a proper and more accurate simu-lation of air valve behavior in water pipelines. Such knowledge gapsand recommendations for further research are discussed in the paper,with a view toward a more efficient application of air valves.
References
Arregui, F. J., et al. (2003). Air valves dynamic behavior, ITA, UniversitatPolytecnica Valencia, Valencia, Spain.
AWWA (American Water Works Association). (2001). Manual of watersupply practices—M51: Air-release, air-vacuum, and combinationair valves, 1st Ed., Denver.
Bergant, A., Bournaski, E., Arregui, F., and Kruisbrink, A. (2004).“Column separation measurements in a large-scale experimentalapparatus.” Proc., 9th Int. Conf. on Pressure Surge Analysis for Designand Operation, BHR Group, Bedfordshire, U.K., 589–604.
Bergant, A., Kruisbrink, A., andArregui, F. (2012). “Dynamic behaviour of airvalves in a large-scale pipeline apparatus.” J. Mech. Eng., 58(4), 225–237.
Besner, M. C., et al. (2010). “Negative pressures in full-scale distributionsystem: Field investigation, modelling, estimation of intrusion volumesand risk for public health.” Drinking Water Eng. Sci., 3(2), 101–106.
Bianchi, A., Mambretti, S., and Pianta, P. (2007). “Practical formulas forthe dimensioning of air valves.” J. Hydraul. Eng., 10.1061/(ASCE)0733-9429(2007)133:10(1177), 1177–1180.
Cabrera, E., Fuertes, V. S., García-Serra, J., Arregui, F., Gascon, L., andPalau, V. (2003). “Reviewing air valves selection.” Proc., Pumps,Electromechanical Devices and Systems Applied to Urban Water Man-agement, A. A. Balkema, Lisse, Netherlands, 633–640.
Carlos, M., Arregui, F. J., Cabrera, E., and Palau, C. V. (2011). “Under-standing air release through air valves.” J. Hydraul. Eng., 10.1061/(ASCE)HY.1943-7900.0000324, 461–469.
Collins, R. P., Boxall, J. B., Karney, B. W., Brunone, B., and Meniconi, S.(2012). “How severe can transients be after a sudden depressurization?”J. Am. Water Works Assoc., 104(4), E243–E251.
De Martino, G., Fontana, N., and Giugni, M. (2008). “Transient flowcaused by air expulsion through an orifice.” J. Hydraul. Eng., 10.1061/(ASCE)0733-9429(2008)134:9(1395), 1395–1399.
Devine, J. G., and Creasey, J. D. (1997). “Surge suppression: Care at allstages.” J. Chartered Inst. Water Environ. Manage., 11(5), 341–345.
DVGW (German Technical and Scientific Association for Gas and Water).(2005). “Dynamic pressure variations in water supply systems.” W303,Wirtschafts- und Verlagsgesellschaft Gas und Wasser mbH, Bonn,Germany.
Escarameia, M. (2005). Air problems in pipelines, a design manual,HR Wallington, Oxfordshire, U.K.
Espert, V. B., García-Serra, J., and Koelle, E. (2008). “The use of air valvesas protection devices in pressure hydraulic transients.” Proc., SurgeAnalysis-System Design, Simulation, Monitoring and Control, 10th Int.Conf. on Pressure Surges, BHR Group, Bedfordshire, U.K., 183–197.
European Commission. (1997). “European Pressure Equipment Directive.”PED 97/23/EC, Official Journal of the European Communities, L181(1),⟨http://ec.europa.eu/enterprise/sectors/pressure-and-gas/documents/ped/⟩.
Gregory, J., Henry, P. E., and Aaron, N. D. (2006). “Meeting the standard—Air release valve and PCCP rehabilitation.” Proc., Pipeline DivisionSpecialty Conf., ASCE, Reston, VA, 49–56.
Lee, T. S. (1999). “Air influence on hydraulic transients on fluid systemwith air valves.” J. Fluids Eng., 121(3), 646–650.
Lee, T. S., and Leow, L. C. (1999). “Numerical study on the effects of airvalve characteristics on pressure surges during pump trip in pumpingsystems with air entrainment.” Int. J. Numer. Methods Fluids, 29(6),645–655.
Li, G., Baggett, C. C., and Rosario, R. A. (2009). “Air/vacuum valve break-age caused by pressure surges—Analysis and solution.” Proc., WorldEnvironmental and Water Resources Congress: Great Rivers, ASCE,Reston, VA, 112–121.
Lingireddy, S., Wood, D. J., and Zloczower, N. (2004). “Pressure surgesin pipeline systems resulting from air releases.” J. Am. Water WorksAssoc., 96(7), 88–94.
Lubbers, C. L. (2007). “On gas pockets in wastewater pressure mains andtheir effect on hydraulic performance.” Ph.D. thesis, Delft Univ. ofTechnology, Delft, Netherlands.
Lubbers, C. L., and Clemens, F. H. L. R. (2005). “Capacity reductioncaused by air intake at wastewater pumping stations.” Proc., Waterand Wastewater Pumping Stations Conf., BHR Group, Cranfield,U.K., 119–131.
Lucca, Y. F. L. D., Aquino, G. A., and Filho, J. G. D. (2010). “Experimentalapparatus to test air trap valves.” IOP Conf. Series: Earth and Envi-ronmental Science, Vol. 12, Institute of Physics Publishing (IOP),Great Britain.
Martin, C. S., and Lee, N. (2012). “Measurement and rigid column analysisof expulsion of entrapped air from a horizontal pipe with an exit orifice.”Proc., 11th Int. Conf. on Pressure Surges, BHR Group, Cranfield, U.K.,527–542.
Martin, C. S., and Lee, N. H. (2000). “Rapid expulsion of entrapped airthrough an orifice.” Proc., 8th Int. Conf. on Pressure Surges-SafeDesign and Operation of Industrial Pipe Systems, BHR Group,Cranfield, U.K.
McPherson, D. L. (2009). “Air valve sizing and location: A prospective.”Proc., Pipeline Division Int. Conf. Pipelines 2009: Infrastructure’sHidden Assets, Curran Associates, San Diego, 905–919.
McPherson, D. L., and Haeckler, Ch. (2012). “Untangling the mysteries ofair valves.” Proc., Pipeline Division Int. Conf. Pipelines 2012: Innova-tions in Design, Construction, Operations, and Maintenance—DoingMore with Less, R. J. Card and M. K. Kenny, eds., ASCE, Reston,VA, 983–989.
Pothof, I. W. M., and Clemens, F. H. L. R. (2010). “On elongated air pocketsin downward sloping pipes.” J. Hydraul. Res., 48(4), 499–503.
Radulj, D. (2007). “The role of air valves in water distribution systems.”B.Sc. thesis, Univ. of Toronto, Toronto.
Ramezani, L., and Karney, B. (2013). “Water column separation and cavitycollapse for pipelines protected with air vacuum valves: Understandingthe essential wave processes.” Hydraulic Surge and Air Control Symp.Number 1, A.R.I Flow Control Accessories, Kibbutz Kfar Charuv,Israel.
Stephenson, D. (1997). “Effects of air valves and pipework on waterhammer pressures.” J. Transp. Eng., 10.1061/(ASCE)0733-947X(1997)123:2(101), 101–106.
Tchobanoglous, G., and Metcalf, E. (1981). Wastewater engineering: Col-lection and pumping of wastewater, 3rd Ed., McGraw-Hill College,New York.
Thomas, S. (2003). Air management in water distribution systems: A newunderstanding of air transfer, Clear Water Legacy, Burlington, ON,Canada.
Tukker, M., Kooij, K., and Pothof, I. (2013). Hydraulic design and man-agement of wastewater transport systems CAPWAT manual, StichtingDeltares, Netherlands.
Wylie, E. B., and Streeter, V. L. (1993). Fluid transient in systems, Prentice-Hall, Englewood Cliffs, NJ.
Yang, K. L., and Shi, W. X. (2005). “Control of pipeline hydraulictransients in south to north water transfer project.” Shuili Xuebao/J.Hydraul. Eng., 36(10), 1176–1182.
© ASCE 04015017-10 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.
Zhou, F., Hicks, F., and Steffler, P. (2004). “Analysis of effect of air pocketon hydraulic failure of urban drainage infrastructure.” Can. J. Civ. Eng.,31(1), 86–94.
Zhou, F., Hicks, F. E., and Steffler, P. M. (2002a). “Observations of air–water interaction in a rapidly filling horizontal pipe.” J. Hydraul. Eng.,10.1061/(ASCE)0733-9429(2002)128:6(635), 635–639.
Zhou, F., Hicks, F. E., and Steffler, P. M. (2002b). “Transient flow in arapidly filling horizontal pipe containing trapped air.” J. Hydraul.Eng., 10.1061/(ASCE)0733-9429(2002)128:6(625), 625–634.
Zhu, M., Zhang, X., Zhang, Y., and Wang, T. (2006). “Study on water ham-mer prevention in pumping water supply systems by multi-valves.”Proc., Int. Conf. on Hybrid Information Technology, IEE ComputerSociety Press, Los Alamitos, 342–346.
Zloczower, N. (2010). “Reduction of damage and hindrance to wastewatertransmission systems caused by air pockets by deployment of airvalves.” Proc., Pipelines 2010: Climbing New Peaks to InfrastructureReliability—Renew, Rehab, and Reinvest, ASCE, Reston, VA,449–462.
© ASCE 04015017-11 J. Water Resour. Plann. Manage.
J. Water Resour. Plann. Manage.
Dow
nloa
ded
from
asc
elib
rary
.org
by
UN
IVE
RSI
TY
OF
TO
RO
NT
O o
n 03
/26/
15. C
opyr
ight
ASC
E. F
or p
erso
nal u
se o
nly;
all
righ
ts r
eser
ved.