8.40

Water Supply Plant Controls P. Y. KESKAR

(2005)

Reviewed by B. G. Lipták (2005)

INTRODUCTION This section describes some of the control strategies and instrumentation systems that can be used in the treatment of raw water in the preparation of drinking water supplies. The described control systems correspond to applications in which PLC-based controls were used. This is not intended to suggest that a PLC-type of control hardware is preferred for these types of applications, but only that the author’s experience was based on that type of hardware. The last paragraph of this section provides a brief discussion of some of the PLC limitations of the PID loops described in this section. The previous section, Section 8.39, discussed wastewater treatment controls. Because some of the unit processes are utilized in both types of plants, there is some overlap between the coverage of these two sections, and it is recommended to review both. This section will describe the controls of conventional water treatment plants, which perform lime softening, filtration, and chlorine- or ozone-based disinfection. The section will also describe the control of water treatment plants using reverse osmosis or submerged ultrafiltration membrane technologies. Examples of actual treatment plants will also be described by flow diagrams.

CONVENTIONAL WATER TREATMENT PLANTS Figure 8.40a shows a simplified block diagram of the unit processes involved in a typical groundwater treatment plant, where raw water is pumped from wells into the treatment units that perform lime softening. In this step, the lime for softening (LS) and the polymer (PO) that assists in the coagulation are added by metering pumps. The coagulated calcium carbonate and colloids, which were used in the water softening step, settle down as chemical sludge (CHS). This sludge is periodically discharged into the washwater recovery and sludge handling unit process by a blow-down pump station. The softened water (FI in Figure 8.40a) flows into the filter influent channel, where carbon dioxide is added to stabilize it and to bring the alkalinity of the water down to 9.5 pH. The pH-adjusted water is filtered in a series of filtration steps, and the filtered and treated water (FE) is sent to a clear well for transferring into ground storage tanks. A disinfectant

© 2006 by Béla Lipták

chlorine solution (CS) is added in the softening, in the filtration, and again in the clear well steps. A high-service pumping station pumps the water from the ground storage tanks (FW) into the water distribution system. Filters are periodically backwashed and air scoured to clean the filter media. The backwash water from the filters is drained (BWD) to the washwater recovery process. The washwater recovered from the washwater recovery process is recycled (BWR) back to the treatment units, and the sludge (DWS) is pumped to vacuum filters for dewatering and disposal. The sludge thickener also receives sludge (DWS) from the washwater recovery basin. Sludge pumps transfer the sludge from the washwater recovery basin to the sludge thickener (see Figure 8.39cc). The thickener decant is received in the decant pump station, where the water is pumped back to the treatment units, and the sludge is pumped to the vacuum filter for dewatering and disposal. The paragraphs that follow will describe the major unit processes and their associated controls. Filtration Hardware and Controls The clarified and softened water from treatment units enters the filter influent channel common to several filters (FI in Figure 8.40a). Here, the water passes through the filter media for the removal of suspended solids from the water such as silt, clay, colloidal particles, and microorganisms. Filtration removes disinfection-resistant microorganisms and colloidal particles that can cause growth and deposition in piping and can alter the taste of water. Figure 8.40b illustrates a typical filtration process and its related controls. In this configuration, the filter effluent and backwash flows, the filter differential pressure, the influent level, and turbidity are continuously measured, and the effluent flow is continuously throttled. Signals from all valves and instruments are connected to a local control panel that houses a remote I/O rack. On/off sequencing valves receive open/close commands from the PLC, and their status is reported to the PLC. The controls shown in Figure 8.40b can be typical for various numbers of filters. PLC Configuration Figure 8.40c illustrates the configuration of the PLC equipment used in the filtration process that was described in Figure 8.40b. The system consists of a redundant PLC housed

Chemical feed system

2173

Distribution system

8.40 Water Supply Plant Controls

RW

FI

CS

HMP

CS

Treatment units-lime softening

HSF

Filtration

FE

Transfer pumping/ ground storage

BWS

Backwash pumps

FW

High service FW pumping

CO2

Ground water wells

PO

LS

CS

Chlorine disinfection

Raw water Filter influent Filter effluent Finished water Backwash supply Air-low pressure Chemical sludge Backwash return Dewatered sludge Lime slurry Polymer Chlorine solution Hexametaphosphate Hydrofluosilic acid Backwash drain

ALP

CHS BWR

Air-scouring blowers

Washwater recovery & sludge handling

Distribution system

RW: FI: FE: FW: BWS: ALP: CHS: BWR: DWS: LS: PO: CS: HMP: HSF: BWD:

BWD

Carbon dioxide system

DWS

FIG. 8.40a Block diagram of the unit processes of a drinking water supply plant that uses groundwater as its source of water.

in a main control panel, MCP-1. The redundant PLC communicates with eight remote I/O racks over a self-healing fiber-optic data link. Each remote I/O rack is housed in a local control panel (LCP) and receives I/O from a pair of filters. Thus, the system controls a total of 16 filters via eight I/O racks. One local I/O rack is installed in the main panel to receive common signals related to backwash pumps and air-scouring blowers. Logic for sequencing of backwash and filter effluent valve control for each filter is resident in the main redundant PLC. The main PLC communicates with the plant computer system via fiber-optic Ethernet links. This allows the plant computer system HMI to display dynamic displays for all filters and to perform SCADA functions like changing set points. The main PLC panel MCP-1 also includes an industrial PC-based operator interface (OI-P) that allows monitoring and control of all filters as well as balance of the plant unit processes via the Ethernet link to the plant computer system. Each local panel has a local operator interface (OIF) that allows monitoring and local manual control of each pair of filters from the dedicated panel. Other filters can also be monitored from each local operator interface (OI-F).

© 2006 by Béla Lipták

Filter Control and Sequencing The control and backwash sequencing logic associated with Figure 8.40b is typical of PLC-based constant-rate filtration installations, where the level in the filter is controlled by modulating the effluent flow control. The operating states of the filter include “auto filtration” (Modes 1 and 2) and “backwash.” When a filter is in its “Auto Filtration” mode, the PLC controls the opening of the effluent valve (FCV-X-3 in Figure 8.40b) as a function of the level in the filter. If the level is very low, a low–low alarm is actuated and the effluent valve closes. If the level (LIT-X-9) is to be kept between Lmin and Lmax (say, 7.0 and 8.0 ft) and the filter flow rate is adjustable between Fmin and Fmax (say, 1.0 and 2.0 million gal per day, or MGD), in Mode 1 (gap control) the PLC would set the effluent flow as follows: If level is at or below Lmin , the flow is set at Fmin and stays there until the level reaches Lmax. At that point, the PLC switches the flow set point to Fmax, which remains there until the level drops back to Lmin. In Mode 2 (level-proportional), if the actual level is x% above Lmin (for example, at 7.6 ft, the value of x = 60%), the PLC adjusts the flow set point to Fmin(100 + x), or to 1.6 MGD.

Modulate

Backwash flow A A A

A A

A

A

A

Status

Eff. flow A A A A

ZS X-3

E

ZS X-5

E

E

FCV X-4

ZS X-6

FV X-6

BWD FV X-5 Controls typical for 16 filters: X = 1, 2, 3... 16 On-off signals: Analog signals: A

FIG. 8.40b The sensors and valving of one of 16 filters having PLC-based sequencing controls.

FE X-7

BWD

FCV X-3

Common header backwash

Filter #X

Common hdr. drain

FE/BWS

Common header filter effluent

FE

Under drain

ALP Flow streams: • ALP - Air low press. • BWS - Backwash supply • BWD - Backwash drain • FE - Filter effluent • FI - Filter influent

From air scour blowers

FIT X-11

FIT X-7

E

FV X-2

To other filters

BWS ZS X-4

Filter media ALP

Modulate

A A PDIT X-8

E

ALP

Status

Press difference A A

ZS X-1 FI FV X-1

ZS X-2

© 2006 by Béla Lipták

Status

AIT Turb. X-10 AE X-10

A

FI

Open close

A A

A E

Status

Turbidity

Level A

Status

Open close

Open close

Status Common header Influent

Common header-air

LIT X-9 LE X-9

Control and Optimization of Unit Operations

To other filters

Open close

2174

Remote I/O rack (PLC based system)

To other filters

From B/W pumps

8.40 Water Supply Plant Controls

Redundant fiber-optic ethernet datalink to plant computer system

Main control panel MCP-1 LIO-1 Common I/O points P R S B I/O C

Main PLC P C R E S P B N BCM U C B

F.O.

2175

F.O.

Ethernet switch

B/UP PLC

P C R E S P B N BCM U C B

OI-P

F.O. “Self healing” fiber-optic RIO link

LCP-1

LCP-2

LCP-7

F.O. P R S B I/O C

PS: CPU: RBC: ENB: BCM: OI-P: OI-F: RIO:

LCP-8

F.O.

F.O.

F.O.

P R S B I/O C

P R S B I/O C

P R S B I/O C

OI-F

OI-F

OI-F

OI-F

RIO rack-1 Filters 1, 2

RIO rack-2 Filters 3, 4

RIO rack-7 Filters 13, 14

RIO rack-8 Filters 15, 16

Power supply Central processing unit Remote base controller Ethernet card Backup communication module Plant operator interface Filters operator interface Remote operator interface

FIG. 8.40c PLC main and subpanels are connected by “self-healing” fiber-optic links and with redundant FO Ethernet links to the plant computer.

If the effluent valve opening reaches 90% during the auto filtration phase, this initiates an alarm condition, indicating that the pressure available for valve differential is becoming insufficient, probably because the drop across the filter bed is high (PDIT-X-8 in Figure 8.40b). Under such conditions, the operator might initiate a backwash cycle. Backwash Cycle The sequence of the backwash cycle can be initiated either manually by the operator or automatically by the PLC. The PLC controls the operation of a number of filters (for example, 16) and will allow only one filter to be backwashed at a time. Automatic backwash is initiated by the PLC, if the preset limits are exceeded on any of the following process variables (Figure 8.40b): level (LIT-X-9), turbidity (AIT-X-10), pressure drop across filter media bed (PDIT-X-8), or when the total run time reaches a preset limit. Table 8.40d lists the

© 2006 by Béla Lipták

steps that occur in a sequence during the automatic backwashing of a filter.

Transfer Pumping and Ground Storage As shown in Figure 8.40a, the filtered effluent (FE) flows into a clear well, where it is further disinfected by a chlorine solution (CS) until the required residual chlorine concentration is reached in the finished water (FW). For detailed discussions of chlorination controls, refer to Sections 8.31 and 8.39. Transfer pumps are used to pump the finished water from the clear well to the ground storage tanks, before it is sent into the water distribution system. The transfer pumping station consists of both constant- and variable-speed pumps and is operated under level control. For a detailed discussion

2176

Control and Optimization of Unit Operations

TABLE 8.40d Definitions of the Steps in the Auto-Backwash Sequence Step No.

Description

0

Out of service

1

Filtration

2

Washwater recovery basin not at high level

3

Close influent valve (FV-X-1)

4

Open drain valve at Falling Level 1 (FV-X-5)

5

Close effluent valve at Falling Level 2 (FCV-X-3)

6

Wait for preset time

7

Start air scour blower

8

Open air supply valve FV-X-2

9

Air scour at preset airflow for preset time

10

Enable washwater control valve (FCV-X-4)

11

Start backwash pump

12

Backwash at low rate with air scour

13

Open air vent valve at Rising Level 3 and close air supply valve (NOTE 2)

14

Stop air scour blower

15

Backwash at mid-rate for preset time (NOTE 1)

16

Backwash at high-rate for preset time (NOTE 1)

17

Backwash at mid-rate for preset time (NOTE 1)

18

Backwash at low rate for preset time (NOTE 1)

19

Close washwater control valve (FCV-X-4)

20

Stop backwash pump

21

Wait for preset time

22

Close drain valve (FV-X-5)

23

Open influent valve (FV-X-1)

24

Open filter-to-waste valve at Rising Level 4

25

Close filter-to-waste valve after preset time (FV-X-6)

26

Select standby/in filtration mode

27

Filter back in filtration

Notes: 1. Low, medium, and high backwash flow rates are set by a flow control algorithm, which detects the actual measured flow (FITX-11), changes the flow set point as required and modulates the control valve (FCV-X-4) to meet these set point settings. 2. The blowers are positive displacement (PD) blowers; therefore, the vent valve at the blower discharge header must be opened before closing the air scour valve FV-X-2. 3. All valve positions are confirmed via limit switches before advancing to next step.

of pump station control and optimization, refer to Sections 8.34 and 8.35. High-Service Pumping The finished water from the storage tanks is pumped into the distribution system by the high-service pumping station,

© 2006 by Béla Lipták

which consists of both variable- and constant-speed pumps. The flow and residual chlorine concentration in the distribution header is continuously monitored, while the pressure is controlled. The high-service pumps are controlled and sequenced to maintain the pressure in the distribution header within a ±2 psi band around the desired pressure set point. A PID algorithm is provided in the PLC to modulate the variablespeed pumps to maintain this pressure set point. In order to properly design a pumping station, one needs to develop reliable system curves for the process, obtain the characteristic curves of the individual pumps, and identify such operating constraints as the surge lines. Based on such information, static and dynamic models can be developed and used to simulate pump station operation using a variety of alternate pump combinations and strategies, before they are implemented. The key to success is using reliable data to obtain the correct system and pump curves for all possible combinations of load conditions. The goal is to avoid overdesigning, which can cause excessive pump cycling, pressure fluctuation, and waste of pumping energy, yet provide all the capacity that is needed to meet present and future loads. For a detailed discussing of pump station control and optimization, refer to Sections 8.34 and 8.35.

OZONATED WATER TREATMENT The conventional water treatment system shown in Figure 8.40a utilizes disinfection steps using chlorine. Figure 8.40e describes the configuration of the equipment in a water supply plant where ozone is used for disinfection. The plant that is illustrated has two ozone generators, each supplying a common header, which takes the ozone to two ozone contactors. Each contactor supply line serves five ozone diffusers, which mix the ozone with the raw water received from sedimentation basins. The ozonated water effluent (OZW) from the contactors is sent for filtration to the biologically activated carbon (BAC) media filters. The excess ozone (OZA) that accumulates above the water surface in each of the contactors is removed and destructed. Each ozone generator includes a power supply for the generation of high-amplitude, high-frequency voltage, which is applied to the electrodes in the ozone generator, which ionizes the oxygen-rich air (GOX) supplied to the ozone generator. The frequency of the power supply unit output is modulated to vary the rate of ozone production. An ozone production turndown of 10:1 can be achieved by varying the frequency from 1000 to 100 Hz. In another method of ozone production, the frequency is kept constant and the ozone production rate is adjusted by varying the amplitude of the voltage. The ozone production rate can be modulated either in proportion to the raw water flow rate (feedforward) or to maintain the desired ozone

8.40 Water Supply Plant Controls

Nitrogen boost (air) ALP

OZA OZ

Oxygen gas GOX

Chilled water

2177

Ozone generator #1

OZ

Power Power supply unit #1

CWS To chillers

03 contactor #1

OZW

CWR Raw water

Treated water

OZA

To filtration

OZ

GOX

Chilled water

OZ

Power supply unit #2

CWS

To chillers ALP: GOX: CWS: OZ: OZA: OZW: CWR:

Ozone generator #2

CWR

Air low pressure Oxygen gas Cooling water supply Ozone gas Ozonated air Ozonated water Cooling water return

03 contactor #2

03 Destruct

B-1

03 Destruct

B-2

03 Destruct

B-3 Blowers

FIG. 8.40e The configuration of the unit processes in a drinking water supply plant, where ozone is utilized as the disinfectant.

concentration in the water effluent (feedback). In either case, the control signal modulates the power supply inverter circuits. As shown on Figure 8.40e, both the power supply units and the ozone generators generate significant amount of heat. Chillers and cooling water circulation pumps are provided to remove this heat and to cool this equipment to the desired operating temperature.

As illustrated in Figure 8.40f, the oxygen supply to the ozone generator can be under either “pressure” or “flow” control. The operator can select (FY-1-1-2) the output of either the pressure controller (PIC-1-1) or the flow controller (FIC-1-1) to modulate the oxygen supply valve PCV-1-1. Depending on that selection, either the pressure or the flow in the oxygen supply header will be held on set point by the corresponding PID algorithm in the PLC. The set point for the oxygen supply pressure controller is set by the operator.

Ozone Generator Controls Figure 8.40f illustrates how a PLC system might approximate, by a series of simple calculations, the operation of a cascade control loop. In this case, the cascade master is the ozone pressure to the contactors (PIC-1-2) and the cascade slave is the oxygen flow (FIC-1-1) into the ozone generator.

© 2006 by Béla Lipták

Calculating the Slave Flow Controller Set Point In the following six steps, a sequence of simple mathematical calculations will be described. This procedure illustrates how PLC-based systems can approximate the operation of cascade control systems.

2178

O2 Flow S.P. per generator step #6

0−1800 SCFM

0−900 SCFM S.P.

0−900 SCFM

to 03 gen #2 FIC-2-1

S.P.

0−300 MGD 1 To contactor #1 03 flow control, Fig. 8.40l

Total trimmed O2 flow (SCFM) step #5

0−1800 SCFM 0-900 SCFM Total O2 flow S.P. (SCFM) step #3 0−900 SCFM

Note 1 (typ.)

2 To contactor #2 calcs

0−150 MGD

Contactor #1 dosage MG/L ∗ 0−10 MG/L

Contactor #1 OZ flow SP step #2

0−12,500 PPD

Contactor #1 OZ demand PPD step #1

Contactor #1 calcs.

0−12% Contactor #2 calcs. note 2

Compute SCFM trim step #4

OZ Conc.S.P. % wt-OZ ∗

0−12%

1

2 ∗

P

FIT 1-1

PIT 1-1

O2 supply

Power supply unit

Ozone generator #1

OZ

PIT From 1-2 03 gen #2 PIT-2-2

AE 1-1

FE 1-1

Notes: 1. For calculations required in each step, see text. 2. Contactor #2 calculations similar to #1. 3. *Indicates manual data entry. 4. All process values are for illustration only.

PY 1-2

AIT 1-1

Power O2

PCV 1-1

PIC 1-2

S.P.∗

MGD: Million gallons per day R.W.: Raw water MG/L: Milligram per liter PPD: Pounds per day OZ: Ozone

A

On-of signal Analog signal PLC algorithm

Select PIT from lead gen.

A

Z

To OZ gen #2 AIC-2-1

A

AS

AIC 1-1

A

A

FY 1-1-1

A

IA

A

S.P. A

PIC 1-1

A

FIC 1-1

A

FY 1-1-2

A

Mode select press/flow

To contactor #1

To contactor #2

From ozone generator #2

OZ

FIG. 8.40f Ozone concentration is controlled by modulating the frequency of the power supply, while the oxygen feed rate is throttled either on the basis of holding the inlet or the discharge pressure of the generator constant.

© 2006 by Béla Lipták

Control and Optimization of Unit Operations

0−150 MGD R.W. flow per contactor

Plant R.W. flow

8.40 Water Supply Plant Controls

The set point of the oxygen flow controller to the ozone generator is calculated on the basis of raw water flow, ozone dosage requirement, and ozone concentration measured in the effluent as follows: STEP 1:

PPD = RWF × Dosage × 8.34

8.40(1)

where PPD is pounds per day of ozone demand in Contactor 1 RWF is raw water flow per contactor in million gallons per day (mgd) Dosage is the required ozone dosage in mg/liter STEP 2: O3 Flow#1 = [(PPD)/(O3% wt × .083 × 1440)] × 100 8.40(2) where O3 Flow#1 is the O3 flow set point in SCFM for the flow into Contactor 1 O3% wt is the set point for the ozone concentration in percentage by weight Note: The ozone flow S.P. for Contactor 2 is calculated the same way as for Contactor 1. STEP 3: Total O2 Flow = O3 Flow#1 + O3 Flow#2

8.40(3)

where Total O2 Flow is the set point for FIC-1-1 in SCFM. STEP #4: O2 Flow Trim = [(PIC-1-2 output)/100] × Trim Range – (Trim Range/2) 8.40(4) where O2 Flow Trim is the set point trimming signal in SCFM, received from PIC-1-2. Trim Range is (−100 SCFM to +100 SCFM) = 200 SCFM. PIC-1-2 output is the percentage output signal of the ozone generator discharge pressure controller PIC-1-2. This controller algorithm in the PLC detects the ozone pressure of the lead ozone generator (PIT-1-2 or PIT-2-2) and compares that reading to an operator-entered pressure set point. STEP #5: Total Trimmed O2 Flow S.P. = Total O2 Flow (Step #3) + O2 Flow Trim (Step #4) 8.40(5) where All flow set points and trim are in SCFM. STEP #6: O2 Flow S.P. per generator = Total Trimmed O2 Flow S.P. (Step #5)/2

8.40(6)

Ozone Concentration Control Loop AIC-1-1 in Figure 8.40f controls the percentage concentration of ozone by weight in the ozone generator discharge stream. This controller

© 2006 by Béla Lipták

2179

receives its measurement signal from the ozone analyzer AIT1-1 and maintains the ozone concentration at the operatorentered set point (percentage) by modulating the output frequency of the inverter part of the power supply unit. Ozone Flow Distribution Control Figure 8.40g describes the ozone flow distribution control loops. The set point for the ozone flow to Contactor 1 is calculated by using Equation 8.40(2), described in Steps #1 and #2 earlier. This total ozone flow is distributed among five diffuser tubes in Contactor 1. Each ozone feeding tube is provided with its own flow transmitter, controller, and control valve. A predetermined percentage of the total ozone flow is sent as the set point of the flow controller of each ozone diffuser tube. The percentages can be changed by the operator; they are provided in a lookup table, and their total must add up to 100%. Control of Ozone Destruction As was noted in connection with Figure 8.40e, the excess ozone that accumulates above the water surface in the contactors is exhausted through ozone destructors, which heat it and, by the use of a catalyst, reconvert it back to oxygen. Each ozone destructor is provided with a continuously running exhaust blower, which maintains a slightly negative pressure within the ozone contactor. A single pressure controller (PIC-4 in Figure 8.40h) maintains a slight vacuum, which is measured at the outlet of the “lead” contactor (PIT1-4-1 or 2-4-1) and modulates all three vacuum breaker valves (FCVs). In the destructor, the ozone is preheated to a preset temperature and is converted back to oxygen by a catalyst. An SCRbased solid-state controller controls the heater by changing the SCR firing angle to control the amount of heat generated. The ozone destructors must be on and running properly before any ozone gas is generated. Ozone Generator Start-Up Sequencing The PLC logic also includes an ozone generator automatic sequencer, which initiates and carries out the system start-up and shutdown in a safe manner. Certain prerequisite conditions must be satisfied before an automatic start-up or shutdown sequence can be initiated. Prior to initiating the automatic start-up sequence of an ozone generator, the following major prerequisites must be satisfied: 1) At least one of the two oxygen supply valves of the two ozone generators (PCV-1-1 or PCV-2-1 in Figure 8.40f) must be in automatic control. 2) At least one of the two ozone generators must be in automatic and ready for operation. 3) At least one of the two contactor trains must be enabled and ready for operation. 4) Emergency stop has not been actuated. When all of the above prerequisites are satisfied, the “Ready” status is displayed and a run command can be

2180

Notes: Compute contactor #1 OZ flow S.P. step #2

0−900 SCFM

3. ∗Indicates operator entry.

0−300 MGD

Note 2

4. All process values are for illustration only.

P AS

I/P FY 1-3-2 Z

0−10 MG/L

OZ dosage MG/L∗

P AS

FIT 1-3-3 I/P FY 1-3-3 Z

P AS

A

A FIT 1-3-4

I/P FY 1-3-4 Z

A

A FCV 1-3-3

A

A FE 1-3-3

FE 1-3-4 FCV 1-3-4

P AS

FIT 1-3-5 A

FCV 1-3-2

FIC 1-3-5

A

FIC 1-3-4

A FE 1-3-2 A

FIT 1-3-2 A

FCV 1-3-1

A

A FE 1-3-1

A

A

A

A

A A A

FIT 1-3-1 I/P FY 1-3-1 Z

FIC 1-3-3

A

FIC 1-3-2

A

FIC 1-3-1

R.W. flow per contactor

15%

15%

A

30%

10%

A

30%

0−150 MGD

Compute OZ demand PPD step #1

0−12,500 PPD % Of total OZ flow allocation to each FIC per lookup table

From OZ generators see Fig. 8.40f

Plant R.W. flow

OZ conc. S.P. % wt∗

I/P FY 1-3-5 Z

AS

Ozone

Contactor

No. 1

To 03 destruct units see Fig. 8.40h

FIG. 8.40g Ozone flow controls for distributing the gas among five diffuser tubes.

© 2006 by Béla Lipták

A

2. For calculations required in each step. See text.

Control and Optimization of Unit Operations

1. Control loops for contactor #1 shown. Contactor #2 is similar.

FE 1-3-5 FCV 1-3-5

P

Select PIT from lead contactor A

PIC 4

A

A

A

A

I/P

A

A

FY 1-4-1

S.P.

A A

A

A

Air

A

A

A

A PIT 1-4-1

Destructor pre-heater

FCV 1-4-1

TT 1-5-1

A

To vent Catalyst

Blower #1

I/P FY 2-4-1

A

S.P.

A

From OZ contactor #1 Fig. 8.40g

A

A

A

A

Air

Destructor pre-heater

PIT 2-4-1

FCV 2-4-1

TT 2-5-1

To vent Catalyst

S.P.

OZ destruct #3

Z

skid

P

AS

Air

Destructor pre-heater

TIC 3-5-1

Blower #2

I/P FY 3-4-1

FCV 3-4-1

TT 3-5-1

To vent Catalyst

FIG. 8.40h The ozone is preheated and its vacuum pressure is controlled by throttling FCV valves that admit air into the suction of the blowers.

Blowers #3

8.40 Water Supply Plant Controls

From OZ contactor #2 Fig. 8.40g

skid

P

A A

OZ destruct #2

Z

AS TIC 2-5-1

skid

P

AS TIC 1-5-1

OZ destruct #1

Z

2181

© 2006 by Béla Lipták

2182

Control and Optimization of Unit Operations

TABLE 8.40i Steps in the Ozone Generator Start-Up and Shutdown Sequence Step 0 — System off, idling. System ready. System run command issued.

The main redundant PLC communicates with the overall water plant computer system via fiber-optic Ethernet links. The Ethernet communication allows SCADA functions from the central control room.

Step 1 — Start cooling water pump. Step 2 — Start chiller. Step 3 — Start boost air.

REVERSE OSMOSIS-BASED WATER TREATMENT

Step 4 — Start destruct unit and enable controls. Step 5 — Open oxygen valves. Step 6 — Enable ozone flow control loops. Step 7 — Start ozone generator purge. Step 8 — Start generator, enable generator control loops.

A water treatment plant utilizing the reverse osmosis (RO) process includes a raw water pretreatment section, an RO membrane train section, and a post-treatment with transfer pumping subsystem.

Step 9 — Enable contactor flow controls, system running. **Issue system stop command. Step 10 — Stop generator, disable control loops, purge. Step 11 — Stop purge. Step 12 — Disable ozone flow controls. Step 13 — Close oxygen valves, wait, destructor purge. Step 14 — Stop destructor. Step 15 — Stop boost air. Stop 16 — Stop chiller. Step 17 — Stop cooling water pump.

issued. Once a run command is issued, the system sequencer goes through a series of steps to start the system. Automatic shutdown is similarly sequenced. The major start-up and shutdown steps in the sequence are listed in Table 8.40i. As each step is taken in Table 8.40i, the actions of the previous step are verified and confirmed before the current step is initiated. PLC Hardware Figure 8.40j shows the hardware of a typical PLC-based control system for ozone generation and disinfection. The control components include a main control panel (MCP) that houses a redundant PLC, a local I/O rack, and an operator interface (industrial PC). Each generator has a dedicated control panel, OGCP-1 and OGCP-2, housing a PLC, I/O modules, and an operator interface. Each ozone generator is controlled by its PLC with all related sensors connected to the PLC I/O modules. The chillers, contactors, destructors, oxygen flow controls, and overall sequencing is controlled by the redundant PLC part of the main control panel, and the local I/O rack in the main panel receives all signals related to everything else except the ozone generators. The operator interfaces in each of the panels allow local control and monitoring of the system. The main redundant PLCs, the two generator PLCs, and the three operator interfaces communicate over a high-speed peer-to-peer data link.

© 2006 by Béla Lipták

Raw Water and Pretreatment As shown in Figure 8.40k, this subsystem consists of the raw water wells from which the raw water is pumped into the membrane trains, after chemical pretreatment and filtering. Chemical pretreatment consists of the injection of acid for pH control and of scale inhibitor, which prevents scaling inside the membranes. The starting/stopping and the adjustment of the speed of the raw water well pumps is automatically controlled to maintain the cartridge filter effluent pressure (PT-1-2-2 in Figur 8.40k) and a “well pump sequencer.” The following data is manually entered into the pump sequencer: well number, pump number, the order in which pumps should be started, target speed, target flow, maximum flow, and well water conductivity. The real-time data used by the pump sequencer includes the pump status (on/off), actual speeds, flows, and well levels. Based on the well levels, water conductivity and other process parameters, the operator manually selects the target speeds and flows for each pump and enters this data in the pump sequencer, which then becomes part of the PLC logic. The operator also enters the pump-starting priority (1–8) in the sequencer logic. In the automatic mode, the well pumps automatically start to maintain the cartridge filter effluent (CFE) pressure in their predetermined sequence and operate at speeds determined by the pump sequencer. As the pressure rises, the pumps are automatically stopped in reverse order, determined by the sequencer. If a pump is switched into the manual mode, it can be started and stopped manually, and its speed can be set by the operator. Acid and Scale Inhibitor Controls As shown in Figure 8.40k, acid is injected into the raw water, upstream of the mixer and cartridge filters, and after the sulfuric acid has been mixed with the raw water, the resulting pH is measured downstream of the mixer (AIT-1-5). The stroke of the acid metering pump is adjusted as a function of the raw water flow (FT-1-1) using a look-up table. The speed of the acid pump is set by the pH controller, AIC-1-5.

Fiber-optic ethernet data link to plant SCADA system.

Fiber-optic E-net SW

Note 1

B C M

DH+

Main PLC RIO link

OI Note 3

C P U

P S

B C M

DH+

Back up PLC Local I/O rack (note 2) P S

R B C

OI

OI

Ozone generator #1 control panel (OGCP-1) note 4

Notes: 1. Redundant PLC with hot backup. 2. Local I/O rack receives signals from ozone destruct units, chillers & other auxiliary equipment. 3. Operator interface is touch screen, industrial environmentally hardened personal computer. 4. Each ozone generator includes a dedicated PLC communicating with main PLC via DH+.

FIG. 8.40j PLC hardware configuration used for the ozone generation and disinfection processes.

DH+

C P U

I/O

Note 3 Ozone generator #2 control panel (OGCP-2) note 4

8.40 Water Supply Plant Controls

Power supply Central processing unit-PLC Backup module Remote base communication module Operator interface High speed datahiway, peer to peer link

P S

I/O

Note 3

I/O

Main control panel (MCP)

PS: CPU: BCM: RBC: OI: DH+:

C P U

DH+

C P U

P S

DH+

DH+

P S

B/up link

2183

© 2006 by Béla Lipták

2184

Control and Optimization of Unit Operations

Raw water wells PLC/computer function

P-1 thru P-8 P-1

Field instrument

A

Analog signal

HMI: Human–machine interface VSD:

P-8 Operator uses wells management table @ HMI for control decisions

A

A

Variable speed drive Signals to flow/pressure controller FIC 1-1 (Figure 8.40l)

A

A

A

A

A

A

A

A PI 1-2-1

Stroke PDI 1-3

A

Calc ∗

A

Speed VSD

A

A A

FT 1-1

PT 1-2-2

PT 1-2-1

Scale Inhib. P-2-1 P-2-2

PDT 1-3 FI 1-1

Mixer Cartridge filters

A Calc ∗

Stroke

VSD

AI 1-5 Sulfuric acid P-1-1 P-1-2

AIT 1-6 A

Speed

AIT PH 1-5 A

A

AIC PH A 1-5 PID

PH

A COND AIT Turb 1-7 A

A

A

A

To membrane trains (Figure 8.40l)

AI CND AI Turb 1-5 1-7 ∗ Indicates calculation

based on equation and/or look up table

FIG. 8.40k Raw water pretreatment controls.

The scale inhibitor metering pump is controlled to keep the scale inhibitor concentration (mg/l) in the raw water constant. Here, too, the stroke position of the metering pump is adjusted as a function of the raw water flow (FT-1-1) using a look-up table. The speed of the scaling inhibitor pump is determined by the PLC, using Equation 8.40(7):

The conductivity and turbidity of the conditioned water feed to the membrane trains is displayed for the operators. The water is allowed to reach the membranes only, if the water quality is within acceptable limits. Any excursions are alarmed, and in extreme conditions, trains can be automatically shut down.

% speed = [ (F × D × Ki)/(# of Pumps × % Stroke) ] × 100 8.40(7)

R.O. Membrane Train Controls

where F is the raw water flow in mgd D is the desired inhibitor dose in mg/l Ki is a constant; in this case, its numerical value is 0.0092

Figure 8.40l illustrates a membrane train configuration, where there are four trains working in parallel. Conditioned water from the pretreatment system is received through a common header that supplies the membrane feed pumps.

© 2006 by Béla Lipták

8.40 Water Supply Plant Controls

2185

Notes: Feed to waste

1. Initially this loop is used to maintain pre-set flow (S.P.) later in the sequence the controller is switched to pressure control.

Conc. flow S.P Qc Min. Conc. flow limit

2. An internal control loop also maintains a stage 1 permeate flow set point.

FIC 3-Z-1

FV 3-Z-1 A

Override

P

M

FCV 3-Z-4-1 S.P.

CALC Qp1

A

E P

FIC 3-Z-4-1

P

A CFE Pressure

CND AIT 4-4-1

To degasifier Figure 8.40m

FV 3-Z-4-3 A

FCV 2-9

E

P1

Stage 1 train #4 Note 2

A

A F.P. #4

PH AIT 4-4-1

FIT 3-Z-4

M

E E

Raw water flow

A

From Figure 8.40k

A

FIC 1-1

FCV 3-Z-3-1

A

FI 1-1-1

P2

FIT 3-Z-5

C

M

A

PT 3-Z-1

Note 1

C

A

E Stage 2 train #4

CND AI 4-4-1

PH AI 4-4-1

PID

FIT 3-Z-3

Flow or pressure S.P.

A

PID FIC 3-Z-4

AFD AFD-2-4

A Perm. target flow S.P. Qp F. P. #4/train #4

A

Permeate flow

A

PLC/computer function C Field instrument

F. P. #3/train #3

P Same as above

Analog signal

A CFE:

F. P. #2/train #2

Cartridge filter effect

P

A

C Same as above

AFD:

Adjustable frequency drive C

From cartridge filters (Figure 8.40k)

P

F. P. #1/train #1

A

Same as above

Conductivity S.P.

PLC Calculations: 1. Target conc. S.P. QC = Qp (100 − Y)/Y where Y = [Qp /(Qp + Qc)] × 100

CIC 44-1

3. % Blend B= [RW/(RW + P)] × 100 RW= Raw water bypass flow; P = Sum of permeates Operator entries: (a) Target perm. S.P. = Qp (b) Target recovery = Y% (c) Target 1st stage recovery = Y1%

Conductivity signal A To injection well

A

2. Target 1st stage perm. S.P. Qp1 = Y1 (Qp + Qc)/100

A

E

FT 1-11-1

Raw water (RW) FCV 1-11

FIG. 8.40l The control loops of a number of parallel trains of reverse osmosis membrane unit include the pressure and flow controls of the water supply, the individual concentrate and permeate flows, and the bypass flows to waste and for blending.

© 2006 by Béla Lipták

2186

Control and Optimization of Unit Operations

Each variable-speed membrane feed pump feeds conditioned water to the corresponding train. The permeate flow from each train is taken by a common header to a degasifier and, after that, into a clear well. Concentrate from each train is taken by a common header for dumping into injection well. Each train consists of two stages. In Figure 8.40l, the overall permeate and concentrate flows are designated as “P” and “C,” respectively. Stage 1 and Stage 2 permeate flows are designated as P1 and P2, respectively. A common feed-to-waste bypass allows the conditioned water to be sent to waste into an injection well by control valve FCV-2-9 during initial start-up or during normal operation, should a high-pressure condition arise. Another bypass allows raw water to be blended into the permeate by control valve FCV-1-11 under conductivity control. Concentrate and Permeate Flow Controls The target percentage of overall water recovery (Y) can be calculated by Equation 8.40(8). The values of Y, QP, and QCMIN are all entered by the operator into the PLC: Y = [QP /(QP + QC)] × 100

8.40(8)

where Y is the target percentage of overall water recovery QP is the target set point for the overall permeate flow QC is the target set point for the overall concentrate flow QcMIN is the minimum acceptable overall concentrate flow The set point for the overall concentrate flow (QC) is calculated by Equation 8.40(9) as follows: QC = QP (100 − Y)/Y

8.40(9)

where Y is the recovery in percentage (%) If QC