EDI System Report Final

GE Power & Water Water & Process Technologies GE Betz Pte. Ltd. 10, Tanjong Penjuru Crescent Singapore 608973 T (65) 62684366 F (65) 62640763

E-Cell MK3-Mini Stacks Flow Path Restriction/Blocked Contents Issues Goals References Site Observation/Review Conclusion Recommendations Issues 1) Envitech Vietnam, has reported two E-cell mini stacks were losing production in Amway Purify Water generation system. Issue identified as follows: a) EDI 1 stack S/N5400066 complete block on the Dilute chamber flow path. Partial restriction in Concentrate and Electrode chambers. b) EDI 2 stack S/N5400060 experiencing High differential pressure drop in the Dilute/Concentrate/Electrode chambers.

Goals 1. To assist in the root cause analysis and recommendation corrective actions.

References 1. 2. 3. 4. 5. 6. 7.

E-Cell MK3 Operation Manual. Quatest 3 Test report dated 06 Dec 2010 Quatest 3 Test report dated 06 Dec 2010 New system upgrade water treatment Process Flow D iagram Site operation parameter records. E-Cell normalization Sheet. Emerson pH/ORP relationship chart

GE Power & Water

Report By: Poh Lip Ben APAC Operations Leader Date: 08 December December 2010


Site Observation/ Review 1.

RO 1 RO 2 RO 3 RO 4

RO Operating parameters dated 07 Dec 2010. Pressure Before cartridge 28 psi

Pressure After cartridge 20 psi

Feed Pressure

Final Pressure

Permeate flow

Reject flow

Permeate

RO Recovery rate

130 psi

90 psi

8gpm

36 psi

195 psi

190 psi

7 gpm

2.9 us/cm 0.8us/cm

53.3%

40 psi 28 psi

20 psi

105 psi

60 psi

9 gpm

4.5 gpm 3.5 gpm 5 gpm

64.3%

50 psi

50 psi

215 psi

175 psi

4.8 gpm

10.8 us/cm 0.1 us/cm

3.2 gpm

66.7%

60%

2.

Sodium Hypochlorite dosing into softened water sample prior to Softener storage tank is designed for ON/OFF operation. To maintain consistent dosage the correct Sodium Hypochlorite concentration should be in the day tank, dosing pump should be check and routinely calibrated. Routine grab sample and test for Free and Total chlorine in the ta nk to ensure dosage of chlorine.

3.

There is no calibration/verification of ORP sensor to ensure accuracy of online readings. The incorrect measurement of the ORP could result in the poor control of the de-chlorination process. This could result in the feeding of chlorinated water into the RO and downstream process. Chlorine level in excess of 1.6ppm to 0.5ppm was detected before the RO 3 and after RO 4.

4.

The feed pressure on RO 3 has declined from the initial commissioning stage of 14 0 psi to 105 psi and the RO permeate quality is averagely at 10-12 us/cm, up from initial of 25us/cm. In comparison to the commissioning data, it is very likely that the RO 3 has been subjected to Chlorine attack and a declined in membrane performance are being observed.

GE Water & Process Technology

Report By: Poh Lip Ben APAC Operations Leader Date: 08 Dec 2010


5.

st

The water sample taken on the 1 Dec 2010 on the RO/EDI feed are as follows: PW#1

SWT

Before RO#1

IT-1

Before RO#2

Feed ED#1

Hardness CO2 SiO2

< 0.33 1.8 14.9

<0.33 1.8 15.6

2.0 1.8 3.52

<0.33 1.8 4.96

0.51 1.8 0.62

mg/l as CaCO3 mg/l mg/l

Total Clorine (Cl2)

1.6

0

0.0

0

0

mg/l

SWT

Before RO#3

IT-2

Before RO#4

Feed EDI#2

Hardness CO2 SiO2

< 0.33 1.8 14.9

<0.33 1.8 15.5

2.1 1.8 7.77

< 0.33 1.8 5.09

<0.33 1.8 0.18

mg/l as CaCO3 mg/l mg/l

Total Clorine (Cl2)

1.6

1.4

1.5

0.5

0.5

mg/l

PW#2

Water samples taken throughout systems show some variation but on the given samples there appears to be chlorine breakthrough and that would lead to the frequent failures of membranes and EDI stacks. The considerably high level of silica and hardness in the feed water for EDI would mean that the EDI system would require periodic chemical in place cleaning (CIP) in o rder to maintain good performance of the system. This would also prevent the buildup of scale and foulant on the stack. Timely CIP would be a recommended corrective action for recovering a given stack to its baseline performance, failing that the stacks may be irreversibility fouled and the stack would need to be replace. GE Water & Process Technology



6.

The operating data was fed into the E-cell Normalization worksheet (see below) to study the operating profile prior to the stacks issues. It was noted that parameters from 29 Dec 10 onwards generated alarming pressures profile. This would typically mean that the stack is experiencing some form of abnormalities and corrective action need to be taken to prevent damage to the stacks.

dp (Dilute Outlet - Conc Inlet)

Recovery

TEA

E-factor

(ppm as CaCO3)

(psid)

SiO2 Removal (%)

% change in normalized pressure drops Dilut e

Concent rate

Electro de

18.0

67.1%

1.9

14.29

0%

0%

0%

20.0

66.9%

2.0

13.22

-14%

-4%

0%

18.0

67.1%

1.9

14.37

0%

0%

17%

14.0

69.4%

1.9

13.64

-3%

0%

8%

14.0

69.4%

1.9

13.27

-3%

-13%

8%

12.0

68.1%

1.9

13.27

11%

0%

17%

20.0

68.2%

1.8

14.70

0%

-4%

44%

16.0

69.4%

1.8

14.45

-3%

0%

8%

18.0

69.1%

2.0

12.59

-3%

3%

5%

-50.0

67.9%

1.8

14.45

106%

50%

-350%

15.0

67.9%

1.8

14.45

59%

8%

17%

15.0

67.9%

1.8

13.79

59%

8%

17%

15.0

67.9%

1.9

13.56

59%

8%

17%

15.0

66.3%

1.9

13.45

66%

8%

6%

15.0

68.5%

1.9

12.90

63%

0%

-50%

14.0

69.8%

1.7

13.90

74%

-10%

83%

13.0

67.9%

1.7

14.80

59%

8%

-67%

Value within normal operating range.

Value change maybe due to normal operation or system problem. Continue monitoring! Value change greater than expected. Intervention needed. Consult Owner's Manual or technical support!




7.

On 07 Dec 10, GE and Envitech conducted a reverse flush to eliminate the possibility of upstream debris (from piping work) blocking the stack inlet flow channels. The exercise did not yield any improvement on the stack flow. The flow restriction/blockage is unlikely due to foreign debris in the feed water.

8.

On 08 Dec 10, GE and Envitech also carried out a Low and High pH CIP o n EDI 1 to access the effectiveness of CIP on the flow issue. Similarly this CIP did not yield any improvement results.

9.

Below is an illustration of a typical ORP (mV) versus pH relationship. ORP reading can be reflection of the change in the water pH. As such the ORP reading prior to the RO1/RO3 could change due to the variation of the softener water pH. Therefore the Chlorine content in the Softener water may not be correctly reflected in the changes of the ORP (mV) readings.




Conclusions The incorrect ORP reading could lead to the over-dose of bisulphite and under dose of bisulphite. Over-dosing of Bisulphite can cause bio fouling issue in some RO system depending on the species of organism in the water. It would a good idea to optimized the Sodium Bisulphite dosing system to ensure that residual sulphite is maintain between 0.5 ppm to 1.0 ppm. The under-dose of sulphite will cause the incomplete de-chlorination. This would damage the RO membranes and E-Cell stacks. GE Water & Process Technology



E-Cell operating parameters must be closely monitored by the use of E-cell normalization worksheet to help identify abnormality in the system and pre-empt stack problem(s) so that timely corrective action can be taken. E-Cell stacks are sensitive to hardness (<1ppm) and Silica (<1ppm) content in the feed water. The hardness are removed from the raw water source by means of the Softener ion exchange and RO systems, As such softener and RO system must be kept in optimum operating condition so that hardness can be removed from the EDI feed. As the Softener ion exchange system would be able to remove such contaminates to a certain extend, frequent backwash and regeneration should be carried out to ensure that these contaminates do not leech out into downstream systems. RO system is mainly used for bulk demineralization, Silica is being removed by this process and if there is a decline in the RO % rejection this would impact the EDI Feed water quality. As less Silica would be reject by it and the RO permeate would contain a higher concentration of Silica. Coupled with a high raw water feed, the EDI feed water silica c an exceed the feed limit in these instances. The Chlorination process should be pre-set on the dosing pump and operator intervention required when the outlet chlorine level is below or above the required set point of 1ppm. This will ensure that chlorine is not being overdosed/under dose as the chlorine demand of the water changes with respect to the micro-biological constitutes in the water.

The failure of the RO and E-cell system is a combination of poor operational practice, operating personnel lack of water treatment knowledge, the failure to response to process changes and telltale sign of anomalies in the system parameters.

Recommendations

1. The operation team needs to be trained in the process and system control philosophy of the water generation and Purification process. 2. Routine Calibration of Instrumentation and analyzer must be performed thru a Preventive Maintenance program. Eg. ORP sensor must be verify/calibrated monthly. Flow meter /pressure meter, conductivity meter annually and etc.




3. A online Total Chlorine meter (colorimetric type) can be installed on the Softener water supply line prior to the RO to monitor and ensure that chlorine are not over or under fed. 4. Adding an Online pH monitoring on the Softener water supply line would provide valuable data for the better control of the de-chlorination process. A chart can be tabulated based on these data against the total chlorine readings (grab sample). This chart wi ll further provide a better guideline on the chlorine content after de-chlorination with compensation for pH variations. 5. The RO ORP / bisulphite dose controller, it is recommended to program low limit cut off in the ladder logic to ensure that the system would have a minimum bisulphite dosing even when the ORP is lower than set point. 6. RO 3 membrane should be replace so as to maintain optimum RO performance. Ch3eck the st permeate conductivity to in order to identify the change required (typically 1 stage only). 7. Daily grab sample check points to include but not limited to: a) b) c) d) e)

Test for Chlorine on the feed to the RO (should use Total Chlorine Test). Test for Silica in EDI Feed water and Raw water supply (should use Total Silica Test). Test for Hardness in EDI feed water and Raw water supply (should use Total hardness Test). Test for pH in EDI feed water and Raw water supply . Test for CO2 in EDI feed water (good to have).

8. The EDI 1 Stack should be replaced and send to E-Cell Factory for A utopsy. An autopsy would identify the root cause of the flow blockage and restrictions. With the autopsy findings we may be able to perform some corrective action to try and recover EDI 2 performance. 9. Additional Analysis a) System wide water chemistry analysis. Inlet and outlet of all installed equipment. E.g. Softener inlet/outlet. RO feed, permeate, reject and etc. Water analysis should include: Calcium Bicarbonate Magnesium Sulfate Sodium Chloride Potassium Nitrate Strontium Fluoride Barium Phosphate Iron (total, dissolved and ferrous) ORP GE Water & Process Technology



Silica (dissolved) Total dissolved solids Conductivity

SDI (for RO feed only) pH TOC

The above analysis would help in the re-evaluation of the raw water quality and the compliance to the design specifications for the EDI system.



EDI System Report Final

Recommend Documents