Water and Wastewater Engineering

Подождите немного. Документ загружается.

16-18 WATER AND WASTEWATER ENGINEERING

(a)

(c)

(b)

Floc.

tanks

Rapid

mix

Sed.

tanks

Control bldg.

Clearwell

Filters

floc./sec

tanks

Rapid

mix

Chemical

control

building

Filters

Clearwalls

(b)

Filters

Floc./sec

tanks

Floc./sec

tanks

Chemical &

control building

Rapid mix

Clearwells

(c)

Operations

building

Clearwell

Floc.

basins

Sed.

basins

Rapid

mix

Filters

Rapid

mix

Sed.

tanks

Filters

Control bldg.

Clearwell

Rapid mix

basin

Floc.

basins

Sed.

basins

Filters

Clearwell

Operations

building

FIGURE 16-7

B a sic styles of plant layout. Linear ( a ), Compact ( b ), and Campus ( c ). Future plans shown by dotted lines.

DRINKING WATER PLANT PROCESS SELECTION AND INTEGRATION 16-19

layout. These include: (1) easier construction because each treatment process is easily accessible,

(2) less potential for structural damage between different processes because of differential set-

tling, (3) upgrading with new technology between units is relatively easy, (4) if the site is not flat,

there is

a potential savings in earthwork, and (5) an inherent increase in safety is realized because

of separation from chemical spills and/or fire.

The disadvantages of the linear and campus styles are (1) larger plant site requirements, (2) more

yard piping, (3) operators must travel greater distance between each process, an

d (4) construction

costs may be greater.

When space is not a constraint, a linear layout generally allows the maximum flexibility for

expansion. Redundancy and flexibility are enhanced if the units are interconnected in such a way

that the flow through the plant can be shuttled from one treatment train to another a

s shown in

Figure 16-8 . Because chemicals must be delivered to the plant, connection to the transportation

net becomes an integral part of the layout. Likewise, because residuals are generally transported

offsite, the residuals handling system is an integral part of the plant layout.

Kawamura (2000) emphas

izes two points in laying out the plant. The first is to provide a

single application point for all the pretreatment chemicals. The second is to provide a single

chemical feed point to the filter influent for all the filters and use this scheme for the combined

filter water prior to the clearwell. The single point application scheme is recommended because it

sim

plifies the chemical feed system and provides uniform water quality control.

Plant Hydraulics

The plant hydraulics are represented by a drawing that shows the hydraulic grade line across the

treatment plant. The drawing must show the elevations of the walkway (top of the structure), the

water level, and the bottom elevation of each unit process as well as the invert and crown of all

the connecting pipes and the invert of all the channels. An example hydraulic profile is

shown in

Figure 16-9 .

Flash

mixer

Flow split weir box

w/weir bypass gates

Flowmeter

Flocculation tanks

Sedimentation

tanks

Sedimentation

tanks

Filters

Clearwells

Chlorine

Pump

station

Flowmeter

FIGURE 16-8

P r o cess train scheme to improve treatment flexibility.

16-20 WATER AND WASTEWATER ENGINEERING

755

760

750

745

740

735

730

725

720

715

Existing

underground

reservoir

Existing

suction well

(1 of 2 shown)

Existing

high service pumps

(1 of 4 shown)

Top gallet wall

Filtered water

Existing sub-basement

From

Recarbonation

EL 736.50

EL 734.25

Existing filters

(1 of 4 shown)

HWL  737.40

EL 734.90 

EL 725.95

Floor EL 714.42 

New Existing

Filter influent

To distribution

system

755

760

750

745

740

735

730

725

720

715

From wells

Raw water

Settled water

Finished floor

Aerator

Clarifier 1 Clarifier 2

Recarbonation

tank

Tank

drain sump

Top of weir box

HWL  742.67

HWL  741.12

HWL  740.12

Clarifier inlet

EL 727.00

EL 743.42

EL 736.50

721.08

EL 741.71

Top of weir

Top of slab

720.08

Top of slab

Top of weir

EL 740.71

EL 737.53

716.42

HWL  738.53

To filter

influent

Top of

slab

NOTE: High water levels (HWL) calculated at 23,000 m

/d with parallel

flow through clarifiers and one filter in backwash

FIGURE 16-9

T ypical hydraulic grade line for a softening plant.

DRINKING WATER PLANT PROCESS SELECTION AND INTEGRATION 16-21

I deally, the water flows through the plant by gravity after it is pumped to the head end of the

plant. This minimizes the number of pumps to move the water through the plant. The elevation of

the surface of the water as it flows through the plant follows the hydraulic grade line. These ele-

vations are set by the design based on calculations of headloss through the various s

tructures of

the plant. Once the headlosses are known, the elevations of the surface water are set by working

from a selected elevation for the discharge from the rapid sand filter or from a selected elevation

for the influent to the plant. The elevation of the water surface in each process upstream is set to

overcome the headloss in moving the water to the next downstream proc

ess.

Some of the headloss calculations have already been demonstrated (e.g., for pipes, baffle

walls, and the rapid sand filter in Chapters 3, 6, and 11, respectively). Other headlosses that need

to be estimated are the losses in the orifice weir from settling tanks and the losses in channels

leading from one process to another. The losses

in the orifice weir may be estimated with Equa-

tion 6-18, repeated here:

QCAgh

dorifice

 ()2

(16-1)

where Q  flow rate through orifice, m

/ s

 coefficient of discharge

A  area of orifice, m

g  gravity acceleration  9.81 m/s

h  headloss through the orifice, m

The coefficient of discharge varies from 0.60 to 0.80. The estimate of orifice headloss is illus-

trated in Example 16-1.

Example 16-1. Design the orifice for Stillwater’s launders (Example 7-4). Assume the orifices

are 5 cm in diameter and that they are placed 0.30 m on

centers. From Example 7-4, there are six

tanks for the design flow rate of 43,200 m

/ d and each tank has 55.5 m of launder.

Solution:

a . Estimate the number of orifices per tank.

()( )55 5 211. m of launder sides per launder  110. m

At 0.30 m intervals, the number of orifices is

111 0

0 30

370

m/orific e

orifices

b. Determine the flow rate per tank in compatible units with Equation 16-1, i.e. in m

/s.

43 200

6 86 400

00833

m /d

tankss/d

m /

()( )

 ssper tank

c. Estimate the flow rate per orifice.



00833

370

. m /s per tank

orifices per tan

m /s orifice 



225 10

4 3

16-22 WATER AND WASTEWATER ENGINEERING

d. Calculate the area of each orifice.





p()005

1 963 10

3 2

e. Calculate the headloss for one orifice using a coefficient of discharge of 0.60.

h 





2981

225 10

060 196

4 3

()()(.

..m/s

m /s

33 10

0 022 2 2 2

3 2







m or cm or cm

)

⎛

⎝

⎜

⎞

⎠

⎟

Because the orifices are all at the same elevation, this is the headloss for water flowing

out of the sedimentation tank into the launder. A sketch of the cross section of the laun-

der and the water levels in the tank and launder are shown below.

Tank

Launder

Orifice

2 cm

Comment. This is a theoretical estimate. Corrosion or encrustation of the orifice will change

the diameter and, thus, the headloss. A safety factor of 2 in the estimated headloss is not unrea-

sonable. This would yield an estimate of 4 cm headloss.

The headloss in the fall of the water from the orifice to the surface level of the water in the

launder requ

ires an estimate of the depth of water in the launder. This is, in essence, a design of

the launder width and depth. This is an open channel flow problem. The flow in an open channel

may be estimated using Manning’s equation:

AR S

channel



100

2 3 12

()()

(16-2)

where Q

channel

 flow rate in channel, m

/ s

n  Manning’s coefficient, dimensionless

A  cross-sectional area of flow, m

R  hydraulic radius, m

 A/P

P  wetted perimeter, m

S  slope of bottom of the channel, m/m

The coefficient 1.00 has implicit units of m

1/3

/ s. The wetted perimeter is the perimeter where the

water is in contact with walls and floor of the channel. Manning’s n i s taken as 0.012 for finished

concrete and 0.018 for steel. The estimate of the surface water level in a launder is illustrated in

Example 16-2.

DRINKING WATER PLANT PROCESS SELECTION AND INTEGRATION 16-23

Example 16-2. E stimate the cross-sectional dimensions of a launder for Stillwater’s (Examples

7-4 and 16-1) settling tank and estimate the maximum depth of water in the channel. Assume the

launder channel is steel and has a slope of 0.002 m/m and a freeboard of 10 cm above the water

level in the settling tank.

Solution:

a . An iterative solution is req

uired because neither the depth or width of the flow in the

channel is known. The approach used here is to assume a channel width in the range of

nominal values (i.e., 0.30 to 0.60 m) and use a spreadsheet tool such as Solver* to deter-

mine the height of the water for a solution to Manning’s equation.

b. From Example 7-4, the design flow rate is 43,200 m

/ d for six tanks and there are three

launders per tank. The flow rate at a launder exit is then:

channel

m /d

tanks lau nders



43 200

6 3 8

()( )(

66 400

0 02778

s/d

m /s

)



c. Using the sketch below and assuming a channel width of 0.45 m, the cross-sectional area

of the water flow is

Ay ()()045. m

where y is the depth of flow.

0.45 m

d . The wetted perimeter is then

Py045 2. m ()

e . The hydraulic radius is





()()

()

045

045 2

f . For steel, Manning’s n  0.018.

g. With the variables defined, Manning’s equation is

0 0278

100

0 018

045

2 3

m /sm

 [( )( )]

(

)( )

()

y045 2

0 002



⎡

⎣

⎢

⎤

⎦

⎥

*Solver is a tool in Excel

. Other spreadsheets may have a different name for this program.

16-24 WATER AND WASTEWATER ENGINEERING

h . Simplifying in terms of the variable y,

()()

()( )( )

0 0278 0 018

100 045 0 002

.. .

m /s

/22

2 3

0 0248

045

045 2





()

()()

()

m y

⎡

⎣

⎢

⎤

⎦

⎥

i. The right-hand side of the equation was entered in a spreadsheet with a starting value of

y  0.30 and goal value of 0.0248. The Solver solution is y  0.22219 m or 0.22 m.

j. Combining the results from Example 16-1 and the width and depth found here yields a

cross section as shown in the following sketch:

Tank

Launder

4 cm

5 cm

22 cm

k . Using the sketch above, the headloss for the settling tank outlet is then:

4 5514 0 14cm cm cm cm or m .

Comment. This is an approximate solution because it ignores the fact that there will be a back-

water curve in the launder. A more exact solu tion may be found using the techniques described

by Reynolds and Richards (1996).

E xample 16-3 illustrates the construction of the hydra

ulic grade line.

Example 16-3. Estimate the elevations and plot the hydraulic grade line for a small softening

plant given the dimensions and headloss data shown below. The top of the clearwell storage tank

is at an elevation of 180.88 m above mean sea level. The weir controlling the water level in the

clearwell is set 2.00 m below the top of the tank.

Solution:

a . Begin with the top of the tank (elev. 180.88 m). The weir is set 2.00 m below the top of

the clearwell at elevation 178.88 m.

Tank Overall height, m Water depth, m Headloss, m

Upflow solids contact7.50 6.87 0.14

Recarbonation 3.502.50 0.14

Rapid sand filter 4.80

3.50

Water depth is measured from the filter floor, i.e., the bottom of the drainage blocks.

Tank vertical dimensions and headloss data

DRINKING WATER PLANT PROCESS SELECTION AND INTEGRATION 16-25

b. The bottom of the rapid sand filter is level with the weir at elevation 178.88. At the

maximum head on the filter, the water level is 3.50 m above the bottom of the filter box.

This water level is at an elevation of 182.38 m. The filter box is 4.80 m deep, so the top

of the filter box is at an elevation of 178.88  4.80  183.68 m.

c. The headloss across the recarbonation chamber is 0.14 m. The water surface in the re-

carbonation chamber is placed 0.14 m above the maximum water level in the rapid sand

filter. This yields a water elevation of 182.38 m  0.14 m  182.52 m. The overall

height of the recarbonation chamber is 3.50 m and the water depth is 2.50 m. Thu s, the

water surface is 1.00 m below the top of the chamber. This places the top of the chamber

at an elevation of 182.52 m  1.00 m

 183.52 m and the bottom of the chamber at an

elevation of 182.52 m  3.50 m  180.02 m.

d. The headloss across the upflow solids contact basin is 0.14 m. The water surface in the

upflow solids contact basin is placed 0.14 m above the maximum water level in the recar-

bonation chamber. This

yields a water elevation of 182.52 m  0.14 m  182.66 m. The

overall height of the upflow solids contact basin is 7.50 m and the water depth is 6.87 m.

Thus, the water surface is 0.63 m below the top of the upflow solids contact basin. This

places the top of the upflow solids contact basin at an elevation of 182.66 m 

0.63 m  18

3.29 m and the bottom of the chamber at an elevation of 183.29 m 

7.50 m  175.79 m.

e. The profile of the hydraulic grade line is shown in Figure 16-10 .

Comments:

1 . Losses in pipes and conduits were ignored in this example. Although the distances be-

tween process units is generally not very large,

significant losses can occur if the pipes

are too small or if minor losses due to valves (especially those not fully open) are not

taken into consideration.

2. V-notch weirs may be used instead of orifices.

Upflow solids contact basin

182.66

183.29

182.52

183.52

180.02

182.38

180.88

178.88

168.88

183.68

178.88

175.79

Sludge

ClearwellRecarbonation Rapid sand filter

Water from wells

Outlet weir

FIGURE 16-10

H ydraulic grade line for a small softening plant. Elevations are above mean sea level. See Example 16-3 for calculations.

16-26 WATER AND WASTEWATER ENGINEERING

Supervisory Control and Data Acquisition (SCADA)

Until the 1960s, almost all control of the operation of a water treatment plant was by manual

methods either by direct operator intervention or by hard-wired electronic switches. The advent

of modern computing technology has radically changed the design of process instru mentation

and controls.

The current generation of control sy

stem software has evolved from the original distributed

control system (DCS) that was developed for in-plant applications where high-speed networking

was available to the SCADA systems that were originally developed for connection over low-

speed data lines. Over time, the SCADA system and DCS have developed very similar capabili-

ties. In the current literature, the SCADA terminology prevails. It will be used in this

discussion

with the understanding that it applies equally well to DCS.

This discussion is an overview. In keeping with the second canon of the code of ethics, the de-

sign of the instrumentation and control system must be conducted by instrumentation and control

system engineers. Instrum

entation, programming, configuration, and other details of the SCADA

system will not be covered in this discussion. The chapter by Kubel (2005) and publications of

AWWA (1993, 2001) provide a more comprehensive discussion of instrumentation and controls.

For convenience, a glossary of abbreviations and terms used in de

scribing SCADA is pre-

sented in Figure 16-11 .

Driving Forces for Implementing SCADA. There are a number of engineering and business

reasons for implementing a SCADA system (Kubel, 2005):

• A need for improved treatment quality. Modern automation and control allows closer and

more consistent water treatment than is possible with manual operation.

• Tighter regulatory requirements. A s sampling an

d record keeping requirements increase,

significant labor savings can result from adding on-line instruments with automatic data

logging.

• A need to reduce costs. Some cost savings through automation include reduced chemical

usage, reduced equipment wear, and optimization of energy

usage.

Field bus—digital communication to field devices

GUI—graphical user interface

HOA—three-way switch (Hand-Off-Auto) that allows manual (Hand) operation, turns the unit

off, and automatically (Auto) runs the operation

LOI—local operator interface

MCC—motor control center

PDA—pocket computer (personal data aquis

ition)

PI—proportional control algorithm

PID—proportional/integral/derivative control algorithm

PLC—programmable logic controller

RTU—remote terminal unit

SP—set point

UPS—uninterruptible power supply

FIGURE 16-11

Glossary of SCADA terms .

DRINKING WATER PLANT PROCESS SELECTION AND INTEGRATION 16-27

• Improved labor effi ciency. Labor savings are achieved by automating tasks and providing

data that allows operators to anticipate equipment failure and perform preventive mainte-

nance.

• Quicker response to emergencies. Centralized control can reduce the time to make changes

in water production during a crisis.

Hierarchy of Control. Water treatment plants often use a co

mbination of manual, semiauto-

matic, and fully automatic control. The hierarchy generally includes local control and computer

control.

The local control includes, for example, manual controls on motor starters, valve actuators,

and pumps. Operators c an set s

uch things as variable-frequency drive (VFD) motor speeds and

chemical flow rates. It also includes hard-wired interloc ks to function when the computer sys-

tems are not in service and to protect personnel while they work on equipment.

Computer control generally includes all fully au

tomatic operations, manual operation from

the computer keyboard, and optimization by computer algorithms.

Types of Control Algorithms. Some example algorithms include sequencing for start-up and

shut-down of processes and continuous control of process operation. The common algorithms are

feed-forward,

feed-back, proportional, and compound control.

An example of a feed-forward system is the polymer feed shown in Figure 16-12 . The poly-

mer flow is maintained at a fixed ratio to the main flow into the mixing tank. The rate controller

[ F (flow)] continuously computes the required polymer flow based on the main line flow.

An example of the feed-back system is the level controller shown in Figu re 16-13 . The op-

erator enters a level set point. The controller continuously measures the level in the tank. If the

tank level is higher than the set point, the controller adjusts the valve to reduce the flow. If the

tank level is low, the controller opens the valve.

Proportional control adjusts the controlled device as a proportion of the measured process

variable. In contrast to the ratio system, the change in the controlled device rises or falls with

a change in the measured variable according to som e mathematical relationship (e.g., rate of

change, power law, exponential, or cubic spline). The PID offers integration or differentiation

algorithms to refine the

control process. For example, if the process variable changes rapidly,

controller output is reduced by an amount proportional to the rate of change.

Compound control is a combination of techniques. A common combination is feed-forward

plus feed-back. In a highly nonlinear system such as pH

control, feed-forward plus feed-back

and a proportional nonlinear algorithm reduces overshoot and hunting by the feed pumps. Chlo-

rination systems with compound control are also favored to maintain adequate residuals without

over-dosing.

Typical Water Treatment Control Strategies. The following outline provides exam ples of

control of some generic water treat

ment processes (Kubel, 2005).

• Raw water pump control. T ypically the operator sets the desired plant influent flow. This

becomes a flow set point to a feed-back flow controller.

• Coagulation. Coagulant and polymers are fed using flow or flow/turbidity feed-forward

control. Alternatively, feed-back from a streaming current detector is

used.