Abstract
Reverse osmosis (RO) has successfully emerged as a broadly-used commercial water purification technology in recent decades. Nevertheless, RO membrane elements, the core component of this purification process, are frequently subjected to premature degradation and performance deterioration, adversely impacting RO plant operation & maintenance. Membrane autopsy is a well-proven yet under-valued procedure for effectively assessing the condition of membranes and determining the root-cause of performance loss. This paper aims to provide a general methodology utilized commercially to perform membrane autopsy and employ three case studies to explicitly demonstrate the value addition to operators and end-users when applying membrane autopsy in (1) failure analysis and trouble-shooting, (2) operation optimization and routine monitoring, and (3) asset management and maintenance enhancement.
INTRODUCTION
In the last two decades, reverse osmosis (RO) has emerged as a highly recognized and broadly used water purification technology, offering commercial application versatility to process otherwise unusable water into useful water sources meeting multiple water quality objectives (Sharon et al. 2000; Karime et al. 2007). According to a market intelligence study conducted by BCC research (Steven 2015), by 2019 the global market for RO and the associated system components is projected to reach $8.8 billion with a compound annual growth rate (CAGR) of 10.5%, primarily driven by municipal water desalination applications, but also by process water treatment (pharmaceutical, power plant, and micro-electronics etc) and water reuse. At the core of this technological advancement is the RO membrane element, the most critical yet delicate component that performs the physical separation of contaminants from water.
Nevertheless, the RO membrane elements, which often represent one of the most significant investments in the treatment plant, are frequently subjected to premature degradation and performance deterioration, resulting in loss of permeate flux, reduced solutes rejection rates, pronounced differential pressure drop, shortened membrane life and inflated operation costs (Ahmed & Al-Amoudi 2002; Karime et al. 2007; GE Water 2011). Such membrane performance decline is usually caused by chemical/physical damages to the membranes and different types of fouling/scaling of the membranes, and can be further traced to unsuitable process design, inadequate pretreatment provision, inappropriate operation practice and unsatisfactory asset maintenance procedures. When various techniques such as plant performance data analysis, process treatment efficacy examination and general plant audit fail to satisfactorily pinpoint the underlying issues, membrane element autopsy will provide practical insight in deciphering the loss of membrane performance.
Membrane autopsy is a well-proven destructive procedure for assessing the condition of membranes and diagnosing the performance loss of membranes (Darton & Fazell 2001; Stephen et al. 2011). It generally entails the physical inspection of element exterior, element dissection and its interior examination, followed by a suite of physical, chemical and microbiological analytical tests done on the membrane and surface deposits to elucidate various types of membrane defects and deposit identities. This would provide essential and reliable evidence to trouble-shoot, remediate and improve the plant operation (Ted et al. 2004; Xu et al. 2010). However, membrane autopsies have not been sufficiently practiced, despite all the palpable benefits that can be derived therein. Common reasons cited include the destructive nature of membrane autopsy rendering the element unusable thereafter, costs associated performing the autopsy and membrane replacement, the turn-around time for the autopsy to arrest immediate operational upsets etc (Nuria et al. 2013). In recent years, improved analytical techniques have enabled more accurate assessment with faster results delivery at more affordable cost; the membrane elements prices have also dropped significantly, making it acceptable to perform more destructive testing. Hence, the benefits of performing routine membrane autopsy shall be explicated and properly communicated to maximize its value addition to the plant operation and ultimately benefit the end-users, all the water consumers.
The objective of this paper is to describe a crucial set of steps routinely performed for the membrane autopsy, and to highlight, from both the technical and commercial perspectives, the utilization of analytical results gleaned from autopsy to maximize value addition to the plant operation. Three case studies shall be discussed in details to externalize such benefits when applying membrane autopsy in (1) failure analysis and trouble-shooting, (2) operation optimization and routine monitoring, (3) asset management and maintenance enhancement.
MATERIALS AND METHODS
The membrane autopsy methodology utilized in the current work consists of a tailored set of essential tests, techniques and analysis to provide understanding to the loss of membrane performance and evidence-supported technical advice to trouble-shoot and optimize the plant operation.
Membrane sample selection
Selection of representative membranes to undergo autopsy is the crucial first step. When permeate quality is of concern, pressure vessel profiling can be done to select the element showing drastic spike in permeate conductivity for autopsy. For general plant failures, a front-end element or a tail-end element can be selected where the front-end element would typically demonstrate the membrane element state with the most severe fouling while the tail-end element will be most likely associated with serious scaling issues.
Physical inspection
Thorough external and internal (after element dissection) visual observation and documentation of the element physical integrity will be performed. Detailed inspection will be done for itemized measurement and to examine the damages/defects on the exterior encasement and anti-teloscoping devices (ATD), O-ring and brine seal integrity, scrolls and product water tube. Once element is dissected, internal membrane leave counts, visible membrane defects/patches, glue-line evaluation and preliminary surface deposits check will be conducted.
Analytical tests
Fujiwara test
This calorimetric indicative test is used to assess occurrence of oxidative damage, especially caused by exposure to halogen-based disinfectant compounds, on the membrane surface.
Scanning electron microscopy (SEM)-energy-dispersion X-ray spectroscopy (EDX)
SEM provides highly contrasted visual observation of the membrane surface for any breach or configurational modification down to nanometer scale. When coupled with EDX, the technique allows chemical composition analysis with quantification or semi-quantification capability to reveal compositional modification on the membrane surface, deposited foulant identity and visual elemental mapping/distribution at selected locations.
X-ray photoelectron spectroscopy (XPS)
This well-established qualitative and quantitative surface chemistry and material analytical tool helps to measure and disclose key information including elemental composition and changes, empirical formula, chemical state and electronic state of the elements that exist on the material surface.
Fourier transform infrared spectroscopy (FTIR)
This technique is used to provide valuable insights associated with the chemical structure of membrane or the surface deposit. When combined with Attenuated Total Reflectance (ATR), it can postulate the presence/absence of chemical functional groups and change in the chemical bond behavior on the membrane surface through the shifting of characteristic absorbance peaks.
Weight loss on ignition (WLOI)
This widely accepted test offers information regarding the moisture content, gauges the organic/inorganic distribution of surface deposits and estimates the major foulant types contributing to membrane system performance decline.
Micro-biological count (TMC)
This provides an indication of biological presences in the surface deposit to assess the extent of microbial contamination.
Technical review & recommendation
The results of the inspection and analytical tests will be compiled and analyzed to provide an overall summary of the membrane autopsy observations and review of possible causes of membrane system performance decline. Recommendations shall be provided, aiming at arresting potential plant upsets, streamlining operational efficiency, reducing membrane fouling and extending membrane system lifetime.
A flow chart describing the essential membrane autopsy methodology and major steps is summarized in Figure 1 below.
CASE STUDIES RESULTS AND DISCUSSION
Case study 1: Application of membrane autopsy in failure mode analysis and trouble shooting
A water treatment plant in South-East Asia was troubled with poor RO membrane system performance, i.e. extensive membrane fouling, high salt passage, low permeate throughput and overly frequent membrane replacement. The plant operator utilized membrane autopsy to understand the underlying issues for the membrane system failure. Selected images and analysis results are presented in Figure 2 to highlight the issues faced by the plant.
Membrane autopsy documented photos for (a) element ATD and scroll conditions, (b) membrane leave examination, (c) SEM image for the membrane surface abrasive damages, (d) SEM image for membrane surface defects with fouling deposits, and (e) EDX profile for membrane surface foulant.
Membrane autopsy documented photos for (a) element ATD and scroll conditions, (b) membrane leave examination, (c) SEM image for the membrane surface abrasive damages, (d) SEM image for membrane surface defects with fouling deposits, and (e) EDX profile for membrane surface foulant.
The physical inspection revealed heavy membrane fouling, indicating possible pre-treatment process failure and unravelling the reason for low throughput and high replacement rate. The magnified SEM images clearly depicted the abrasive damages on the membrane surface inflicted by fouling deposits, explaining the causes to poor permeate quality. Foulant analysis using WLOI discovered that the blackish/brown deposits contained chiefly inorganic material (i.e. 86%), which was further corroborated by other advanced technique such as EDX-RF, see Table 1, showing that the deposit constituted high content of inorganic manganese and iron.
EDX-RF analysis of the fouling deposit
Element . | Percentage (wt%) . |
---|---|
Manganese, Mn | 84.84 |
Iron, Fe | 8.78 |
Calcium, Ca | 1.92 |
Sulphur, S | 1.10 |
Potassium, K | 1.09 |
Copper, Cu | 0.75 |
Cobolt, Co | 0.71 |
Phosphorus, P | 0.38 |
Strontium, Sr | 0.35 |
Molybdenum, Mo | 0.07 |
Zirconium, Zr | 0.01 |
Element . | Percentage (wt%) . |
---|---|
Manganese, Mn | 84.84 |
Iron, Fe | 8.78 |
Calcium, Ca | 1.92 |
Sulphur, S | 1.10 |
Potassium, K | 1.09 |
Copper, Cu | 0.75 |
Cobolt, Co | 0.71 |
Phosphorus, P | 0.38 |
Strontium, Sr | 0.35 |
Molybdenum, Mo | 0.07 |
Zirconium, Zr | 0.01 |
Supported with such evidence gathered from membrane autopsy, a plant audit was followed up to verify the autopsy findings and postulate the root-causes of failure, and tailor the technical advice and recommendations accordingly. Feed water source was found to be highly anoxic and containing high levels of colloidal and dissolved forms of manganese and iron. The pre-treatment settling stage and media filter was found to be undersized, resulting in insufficient colloids/particulates removal. The water feed tank before RO was found exposed to ambient environment, introducing air which oxidized dissolved metals into insoluble precipitates before RO. The cartridge filters were improperly sized, selected and maintained, unable to offer the functional protection to the downstream RO membrane process. Some of the recommendation rendered to the plant owner included retrofitting the pre-treatment stage with additional settling tanks and media filters; replacing the filtering media with alternatives having appropriate filtration rating; enclosing the feedwater tank to minimize the exposure to open environment; replacing cartridge filters to have the more suitable rating and capacity. With the recommended process modification and amendment implementation, fouling issue has been significantly alleviated and the CIP and replacement frequencies have been much reduced. This is translated into enormous amount of 1–1.5 million USD annual saving just for the RO membrane replacement. Recent RO data has also indicated noticeable improvement in plant performance in terms of permeate quantity and quality, and far less process disruption. All these are further rendered to the plant owner/operators as operational optimization benefits.
In this case, membrane autopsy has provided the forensic membrane condition assessment and foulant analysis which were used as the basis to identify the root causes of the plant performance failure, so that the appropriate corrective remediation/trouble-shooting advices can be implemented to pinpoint the exact issues and effectively arrest the operational upsets.
Case study 2: Application of membrane autopsy in plant operation routine monitoring and performance optimization
The O&M team of a large-scale seawater reverse osmosis (SWRO) plant in the middle east requires technical advisory services in a bid to optimize the plant operational performance and streamline the O&M procedure and practice. Membrane autopsy was performed with detailed examinations and analysis on the used element to gauge the element integrity and understand the foulant composition in a comprehensive manner. This provided the critical supplementary evidence and reliable information to support insightful analysis and specific recommendation to the plant operation & maintenance practice. Selected findings are highlighted in Figure 3 shown below.
Membrane autopsy highlighted findings including (a) membrane surface wrinkle/crease lines, (b) SEM image zooming-in on the cracks along the crease, (c) foulant depositing on the membrane surface, (d) WLOI at 550 °C revealing organic content of the foulant, (e) curve-fitting for foulant analysis using XPS, and (f) FTIR-ATR profile of the foulant.
Membrane autopsy highlighted findings including (a) membrane surface wrinkle/crease lines, (b) SEM image zooming-in on the cracks along the crease, (c) foulant depositing on the membrane surface, (d) WLOI at 550 °C revealing organic content of the foulant, (e) curve-fitting for foulant analysis using XPS, and (f) FTIR-ATR profile of the foulant.
Close examination of the membrane surface underscored the presence of wrinkle-like crease lines on the membrane surface, especially along the membrane glue-line where even minimal movement would be prohibited. This lead to the speculation that in-plane compression of membrane surface over the years had reached critical state, which then resulted in membrane out-of-plane deformation creating crease/wrinkles and tearing damages on the surface. For the foulant deposit, the basic analysis using WLOI revealed that high moisture and high organic contents. Combined with analysis from more advanced techniques such as FTIR -ATR, and XPS in elucidating the empirical bonding structures and associated chemical environment, it was suspected that pre-treatment was inadequate to remove the natural organic materials (NOMs) present in the feed water source (Shon et al. 2007).
Backed with such evidence-supported analysis, rational deduction-based recommendations on pre-treatment operation, cleaning and replacement practices were provided. For example, it has been recommended to the plant O&M team to adhere strictly to proper RO train start-up/shut-off procedure with pressurization of the RO system done at controlled rates. Feed pressure and differential pressure ΔP build-up across the element should also be monitored carefully as the key indicators to initiate CIP and replacement to prevent over-pressurization and membrane deformation. Additional pre-treatment such as DAF and optimizing coagulation dosage should be implemented to more effectively remove NOMs from the feed. With implementation of the recommendations still in progress, the O&M team are already witnessing improved permeate production and extended duration in between CIPs. Membrane autopsy would be performed regularly to document and profile the performance enhancement.
Thus, for this work, membrane autopsy has offered a means to track the RO system performance and identify the operational short-comings with improvement measures to upgrade the plant O&M practice and optimize the overall plant performance (Pontié et al. 2007).
Case study 3: Application of membrane autopsy in chemical selection and asset management enhancement
Membrane autopsy can prove useful when evaluating the efficacy and potential impact of chemicals on the membrane; in this case, a SWRO plant in Indonesia required the service that a rejuvenating chemical has been assessed in restoring/improving the performance of used membrane elements with poor solute rejection. The approach entailed an initial used membrane element inspection and performance evaluation, followed by chemical treatment and subsequently performance re-evaluation. Membrane autopsy was performed as the final step to verify the impact of chemical treatment. Selected highlights are illustrated in Figure 4 to underscore the value addition of membrane autopsy.
Chemical evaluation methodology highlights including (a) performance evaluation of membrane element, (b) probing profile before and after chemical treatment, (c) membrane inspection during autopsy, and (d) SEM image zooming in on the abrasive cracks remaining on the membrane.
Chemical evaluation methodology highlights including (a) performance evaluation of membrane element, (b) probing profile before and after chemical treatment, (c) membrane inspection during autopsy, and (d) SEM image zooming in on the abrasive cracks remaining on the membrane.
The performance comparative evaluation gave an empirical observation regarding the effectiveness of chemical treatment. In this case, probing profile of the element has demonstrated only a slight improvement in the solute rejection of the membrane after the chemical treatment. Membrane autopsy was performed on the used element and SEM images has explicitly shown that the treatment chemical was unable to repair/seal up the large abrasive damages found on the membrane. It was suggested that the chemical treatment procedure shall be modified to enhance the repairing efficacy, or an alternative chemical should be proposed and adopted for the purpose.
For this case, membrane autopsy has served as a verifying step allowing a direct measure of the effectiveness of existing/proposed RO membrane chemical treatment programs. The same methodology is currently in use to aid the plant operators in selecting effective cleaning chemicals for specifically fouled membrane and assist in designing improved cleaning/replacement strategies to achieve asset management enhancement.
CONCLUDING REMARK
Membrane autopsy is a valuable tool to assess the condition of membranes and offer insights elucidating the underlying issues causing membrane system performance loss. It has the flexibility of being a routine stand-alone technique, and it also can work synergistically with other procedures such as plant audit and performance data analysis to pinpoint the system issues and enhance the plant O&M in different scenarios including failure analysis and trouble-shooting for system malfunction, operation optimization and routine monitoring, and asset management enhancement. The utilization of membrane autopsy in a systematic and skillful manner will fully capitalize its value addition to the plant operation & maintenance and ultimately prove to be beneficial to all the water consumers.