OUTLINE OF THE PROJECT

1. STUDY POPULATION ESTIMATE future application of compound interest formula: Pn

= Po (1 + i) n where Po = population note on a generic year Pn = population aft
er n years from the year covered by Po = the rate population growth constant for
the n years of expected life for the aqueduct (provided by PRG), n = number of
years over which you want to know the population Pn. 2. Choice of route network
study is conducted on a plane with dimensions of the center (1: 5000) in which t
hey scored the distribution of population, the extent and character of different
areas. It defines the urban area (the main cities, villages, etc.). Aqueduct co
ncerned on the basis of the Plan or other planning instruments indicate the exte
nt and type of settlements and productive future (Fig. 1).

Figure 1. Diagram of distribution center.

Examination of the altimeter of the areas to be served, hence the decision to ad

opt a single distribution network or different networks with different distribut
ion piezometric level. Indeed, in centers with considerable variations in altitu
de, so that in case more or less extended the network far exceed the 70 m drop i
n hours of consuming less, it is appropriate to consider the creation of more in
dependent distribution networks, each which serves an area with variations in he
ight contained (Fig. 2).

Figure 2. Division into independent networks each of which "dominates" urban are
a

The position of the reservoir area should be broad network that within hours of
consumption and thus minimum tank at the highest level, you have a load not exce
eding 70 m above street level (you can actually believe that in this case the ne
twork we have the hydrostatic conditions), and highest level of consumption, and
therefore with the minimum level in the tank, it has a share piezometric higher
at all points of the network, not less than 10 m beyond the top of buildings. E
ven fluctuations in network load caused by variation in water demand throughout
the day, must be contained within the 15-20 m water column, and this is the regu
larity of water supply service, either to prevent the rapid loss of elasticity o
f the rubber seals of the joints of the pipes of the network, resulting in sharp
increase of water loss (Fig. 3).

Figure 3. Piezometric time of increased consumption and lower consumption of tim

e for networking with tank tested.

Shares and fixed position of the tank, it passes the line of the network, whose
pattern usually can be closed mesh (rings) or branches. The first type is prefer
able because it allows a more flexible service, helps to limit the swings of loa
ds, reduces the areas to be excluded in case of failure, allows the circulation
of water in more ways and more effective service in case of fire. The pipeline r
oute follows the streets, tracking is done considering a hierarchy of pipe: ie f
irst defining the network of

main pipes, also for the provision of fire doors, then the pipes are of secondar
y importance.

3. Trace the size of the network distribution network, we determine the flow of
users, distributed and / or focus, that every body should provide the various op
erating conditions, particularly under conditions of maximum consumption (ie, th
e peak day flow maximum consumption). Therefore, it identifies any conduct for t
he area to be served by the method of bisecting lines using a dimensioned scale
1: 5,000 and taking into account the shape of the city center. Given each area,
we examine which areas of the PRG it includes, and then multiplying the size of
the area on an area for manufacturability index, you get the volume to be constr
ucted to serve in any trunk. On a statistical basis, the literature shows that t
he volume is built can be traced back to the population, requiring that every re
sident is required to manufacture 100 m3. In summary, the design takes into acco
unt time for a network consisting of a number of links, and determines the users
that each section shall serve as follows: AREA area (m 2) × IF (M3/m2) = built
VOLUME (m3) built VOLUME (m3) / 100 = m3 population based on the data obtained w
e calculate the average annual flow Qm to be supplied to different areas, taking
an envelope of water per inhabitant, expressed in the / ab.× day reported the p
lan to restructure the Puglia Region: QM = (P × d) / 86400 (l / s) where: P = po
pulation served in each log d = daily water allocation

and then measuring the maximum to be paid during rush hour: Q = h max × Cp Qm ha
ving evaluated the coefficient bit with the formula of Gibbs: Qh max / Qm = 5.0
/ P 1 / 6 where P is the population served in thousands of inhabitants. This ran
ge is assumed to be uniformly distributed along each pipeline. CALCULATION OF DI
AMETERS order to make a preliminary sizing of the pipes need to know what the fl
ow circulating in the individual branches of the network. And 'know that the des
ign problem, namely the determination of the diameter of a closed system is not
direct or immediate solution: they are in fact known prior to and extent of flow
in each section. Usually proceeds in the design phase, cutting the cracks close
d and following the appropriate points in the case of an open system: that is, a
ssuming that the sections where you cut the network correspond to minimum piezom
etric and, therefore, that in them flow is zero. Breaks, ie the so-called "neutr
al points" (junction points of water) are for link. The sizing of the pipes is t
hen achieved by taking the series involving commercial diameters, calculated for
the flow, speeds between 0.5 and 2 m / s (preferably around 1 m / s), with the
warning not to get never below 100 mm Ø = not to compromise the proper functioni
ng of the network in situations other than the budgeted provision (fires, etc.).
.

In determining the diameter must be careful to choose a few diameters, so that t

he managing body is facilitated in the storage of pipes, special parts and equip
ment handling, and maximize uniform diameters of key links in which the flow con
veyed may differ from those determined in this calculation to a first approximat
ion. Except in special cases, in the case of pressure pipes for water supply are
neglected localized losses (curves, changes in diameter, etc..) Taking into acc
ount only the losses continued, because the length of the pipeline is generally
very high and losses localized (except possibly deliberately caused by regulatin
g valves) can be neglected. The calculation of the diameter must be conducted in
the case of tubes used, as the worst operating condition occurs over time, when
you have a natural increase of roughness, and hence an increase in pressure dro
p, the risk that in this condition The minimum load available is no longer guara
nteed. Assumed then the diameters, you can calculate the pressure drop with Darc
y's formula: LQ ΔY = b * 2 / Dμ where: Q = flow * dummy evenly distributed (m3 /
s), L = length of the trunk network (m ) D = diameter of the trunk network (m)
b, μ = parameters related to the type of pipe selected. For simplicity of calcul
ation, the courses are considered uniformly distributed along each log, using as
a scale for calculating a notional capacity of:

Q * = Qd = qoute qoute + 0.55 + 0.55 (Qin - qoute) = 0.55 + 0.45 qoute with Qin
Q *= scale fictitious Qin = inlet flow in the trunk; qoute = flow emerging from
trunk Qd = flow evenly distributed. CHECK HYDRAULIC NETWORK In meshed closed cir
culating flow are not known and it is therefore necessary in order to determine
a preliminary calculation. Moreover, the verification of network nodes is done b
y imagining concentrated delivery of dishes. So that the network is balanced bot
h in terms of flow rates (tax already provided initially to determine the flow h
ead in different nodes), but also loads piezometric is necessary to adopt a meth
od of calculation to be satisfied that the equations continuity of flow and load
s: Σi (qij ± ± Qj) = 0 (1)

where qij is the flow circulating in the branches that flow into node j, Qj is t
 he total flow of external node, and the sign depends on convention adopted (eg,
the positive flow entering the node); Σi (± Δhi) = Σi (± rik Qik Qik) = 0 (2
)

Qik is where the flow circulating in the branches belonging to the mesh k, rik i
s the constant value characteristic of each branch of resistance in the formula
chosen, and the sign depends on the convention used for the sense of water circu
lation in the mesh (eg , positive if clockwise).

The solution of the system of equations (1) and (2) is not simple, is the high n
umber of equations that comprise it, is the fact that (2) are not linear. Equati
ons (1) and (2) are independent,in a mesh network closed flat, number of m + n
- 1, m being the number of meshes and n the number of nodes. The unknowns are th
e flow circulating in the branches, which are equal in number am + n - 1. We the
refore prefer to use iterative methods, including one of the most traditional is
the Hardy-Cross method. 4. METHOD OF BALANCING THE CRO method consists in a f
irst attempt q'i distribution of circulating flow (flow head) that is consistent
(ie satisfies the system of equations (1) continuity at the nodes), and to oper
ate subsequent corrections until you get to a solution of equations (2) sufficie
ntly approximated, namely to verify the consistency of loads. It then assumes th
at the distribution takes place only at the nodes, as shown in Figure 4.

Figure 4. cheme of a distribution network.

¡ ¡ ¡

Initial data are shown in the table.

JER EY trunks (m) L (m) DATA INPUT bμ ki = bL / Dμ IQ (m3 / s)

The method is applicable to all sections of each mesh. You know:

The course attempts IQ (m3 / s) [end doors Qin] The diameter of each log D (m) [
is obtained from the formula D = (4 Qin / π v) 1 / 2 by choosing the internal di
ameter, closer among those rovided by commercial manufacturers of i es];

The length of each log L (m) The characteristics of roughness (b and μ).

It 'also notes the share piezometric tank (load assigned to node A in the networ
k outlined in Figure 4). The flow Δqk correction of each mesh to calculate the l
oad balancing is provided by: Δqk = - Σ (± rik q'ik q'ik) / 2Σrik q'ik C
orrecting the flow q'i with flow Δqk displayed, with the wisdom to apply common
mesh sides adjacent the algebraic sum of the corrections calculated separately,
repeats for the number of times needed to achieve the desired accuracy (in pract
ice until the sum losses on each link is less than 0.50 m in absolute value, or
until the extent of correction is less than 0.1 l / s for all the loops).

Iteration to iteration occurs also moving the neutral point which converges too,
hovering around its starting position, parallel to the convergence of calculati
on. The 1st iteration provides:
Q (m / s)
The 3

k Q

The

The

1 ª ITERATE ΔY = KI QI QI (m)
 ΔQI

QI = QII + ΔQI

The pressure drop on the trunks of each mesh and then the sum on each jersey [ΔY
= KI QI QI (m)];

The course correction ΔQI of each mesh, the correct ports = QI + QII ΔQI to use
the 2nd iteration.

At the end of the iteration the neutral points of each mesh undergo a shift, as
the incoming flow at the nodes of the network were correct. Therefore vary the l
engths of the portions which are divided into the trunks of the network containi
ng the neutral points and, consequently, the corresponding value of k.
CHANGE OF NEUTRAL POINT: 1st ITERATE MESH DISTANCE FROM node LEFT SHIFT TO THE L
EFT Xsx (m) ΔX (m)

The results of the 2nd iteration are:

QII LII KII KII QII 2nd ITERATE KII QII ΔQII QIII QII QII = + ΔQII

Occur, as in the 1st iteration, the value of the sum of losses and magnitude of
correction and the shift of the neutral point on each link. Thus correcting the
flow with ΔQII you make the 3rd iteration. It repeats for as many times as neces
sarytosatisfythefollowingconditiononeachlink: Σ ΔY ≤0.50m,or ΔQn≤ 0.1 l / s

INSPECTION 1) Check the daytime After the Cross, the final load values are those
which meet the matching and then are ultimately to be taken. You can then calcu
late the ΔY for each trunk, fixing a single share, such as grouse, and drawing b
elow all others. Since these loads were obtained taking into account the capacit
ies that need to be distributed to the people at rush hour, or the maximum consu
mption, they are the minimum possible. Then subtracting the portion of the plan
year of each node of the load supplied, we can obtain the available load day. Sh
ould not exceed 50-60 m. 2) Verify that night there was placed in the toughest c
onditions that we have assumed that there is no consumption at night and then th
e tank is full. That is a condition of hydrostatic load and then load available
at nights is the difference between the highest and the reservoir portion of the
plan year of the different nodes. Must not exceed 70 m approx.3) Verification
of speed are calculated speed trunks with final flow values and verify that they
are less than 2 m / s threshold taken to avoid excessive wear, but not exceedin
g 0.5 m / s to avoid possible sedimentation of suspended solids.

TRUNK NODE

ΔY

LOAD TEST FEES FEES Piezometer pc (M) (m) Hsmin

AVAILABLE LOADS (m) OF DAY-NIGHT Hsmin Hsmax zA-zA

Hsmax

ZA

From node A, which is assigned the hs share, and notes the losses on each tree,