Membrane Autopsy Techniques

Introduction

Many communities rely on membranes to remove pathogens from drinking water. The resilience of water recycling plants that utilise membranes is defined as the ability to meet a specified pathogen concentration log reduction in routine, as well as under unexpected circumstances. It is also an essential factor in ensuring continuous process throughput whilst remaining compliant with strict water discharge guidelines whilst ensuring public health and safety. Given the importance of meeting and maintaining both throughput and quality targets, evaluation of a treatment plant’s resilience is highly necessary and should be performed during its design stages as well as throughout its operational lifetime.

Resilience, however, has traditionally been difficult to evaluate and quantify, thus has largely been ignored leading to the tendency towards a “belt and braces” approach where multiple redundancies are put in place. This consequently results in an over-engineered design that requires a higher capital and operational expenditure over the asset’s lifetime. Despite the common adoption of this traditional approach, the Global Infectious Diseases and Epidemiology Network (GIDEON) database still catalogued 2500 confirmed pathogenic outbreaks from 2003 to 2013. Further investigation revealed that 60% of the causes of failure stemmed from inadequate process design and poor asset management. To date, no standard resilience modelling method has been developed for the water treatment industry therefore, highlighting the need and importance of resilience modelling in this industry.

Highlights

Able to model resilience of water recycling systems with accuracy and consistency Resilience modelling can predict and improve asset performance and reliability Assess individual asset criticality, identify process bottlenecks and weak spots Resilience model provides insights to specific failure modes and resultant effects Compare resilience and benefits of alternative designs and operating strategies

Purpose

OPTAGON Model
Asset Resilience

Resilience is a system’s ability to maintain routine function even under unexpected circumstances.It is an essential factor in ensuring continuous process throughput whilst remaining compliant with strict water discharge guidelines.
Resilience modelling tools have been widely used in the Petrochemical, Oil and Gas, and Aviation industries to model process reliability and safety over the last 15 years.
No standard resilience modelling method has been developed for a potable reuse scheme.

Study used a resilience modelling tool from the Oil and Gas industry, GL Noble Denton’s (GLND) OPTAGON Simulation Package
OPTAGON is GLND’s Monte Carlo-based Reliability, Availability and Maintainability (RAM) simulation tool which is capable of modelling the performance of asset.
With user-variated real-time data, OPTAGON is able to accurately predict equipment failure and system resilience.

Objective

Develop a mechanical resilience model for dual membrane plants (MF/UF + RO) using data from large scale (>10 MLD) plants with long operating history (7-10 yrs)
Develop “What-if” scenarios for resilience model’s sensitivity based on confirmed cases of drinking water plant failure resulting in pathogen infection
Quantify process resilience and predict process equipment failure using resilience model.

Common Failures in Drinking Water Systems

GIDEON database catalogued >2000 confirmed pathogenic outbreaks from 2003 to 2013.
30% of the outbreaks were associated with protozoan parasites.
The most common type of failure was an incident in the catchment area in conjunction with an inadequate process design.
Second most common type of failure occurred in the distribution system followed by an inadequate management framework and operational error.
Poor asset management and failures highlight the need and importance of resilience modelling in the water industry.

Supporting Evidence

Modelling Process

Data Sourcing and Collection

Equipment failure and performance data is sourced from 7 water recycling plants worldwide.
Relevant information is collected from a wide array of data sources.

Data Analysis and Mapping

Cataloged equipment data is sorted and mapped according to process equipment specified in the model (Reference Plant).
Equipment arranged with design and operational capacities based on functional location.
Operation & Maintenance (O&M) Manuals provide vital information on equipment availability.
MTBF and MTTR are also calculated if not previously provided.
Equipment criticality is determined based on failure and maintenance data.

Resilience Modelling and Sensitivity Analysis

Mapped data becomes input variables for OPTAGON to model asset’s mechanical resilience.
Monte Carlo simulations of 10,000 realisations ensure confidence of modelling results.
Results also demonstrate equipment interdependency.
Modelling results would quantify the asset’s overall reliability and resilience.
Sensitivity analysis would further highlight which input variable would have the greatest impact on the system.
“What-if” scenarios would test the robustness of the reference plant and aid with process optimisation.

Resilience = ƒ (Availability, Performance)
Availability = ƒ (Reliability, Maintainability)
Risk = ƒ (Likelihood, Consequence)

Outputs

OPTAGON can model complex water recycling systems with high level of accuracy and consistency.
Modelling results would be able to quantify asset resilience, criticality and risk.
Resilience modelling can predict and improve asset performance throughout asset’s lifespan.
Sensitivity analysis would support asset management decisions and aid in efficiency and profitability.
Reference model can also be used to provide insight to specific failure modes and resulting effects.