5.3 Membrane Integrity Testing & Troubleshooting

Regulatory Framework & EPA Guidance Requirements

Under the Safe Drinking Water Act and the Long Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR) (codified in 40 CFR §141.719(b)), membrane filtration facilities can earn up to 4.0-log (or higher) removal credit for Cryptosporidium and Giardia. However, because membranes consist of thousands of microscopic hollow fibers or delicate spiral envelopes, mechanical failures—such as a single severed hollow fiber—can allow pathogens to bypass the barrier into finished water.

The EPA Membrane Filtration Guidance Manual (MFGM) establishes two mandatory tiers of surveillance:

                      EPA Integrity Verification Architecture
                      
   +---------------------------------------+   +---------------------------------------+
   |     Direct Integrity Testing (DIT)    |   |  Indirect Integrity Monitoring (IIM)  |
   |---------------------------------------|   |---------------------------------------|
   | • Daily Frequency (Every 24 Hours)    |   | • Continuous Real-Time Monitoring     |
   | • Physical Pressure Decay Test (PDT)  |   | • Skid Turbidity (<=0.15 NTU 95% time)|
   | • Resolution: <= 3 µm (Cryptosporidium|   | • Never Exceed 0.30 NTU               |
   | • Quantifies Log Removal Value (LRV)  |   | • Continuous Laser Particle Counting  |
   +---------------------------------------+   +---------------------------------------+

Core Regulatory Criteria for DIT

1. Resolution Criterion: The direct test methodology must be sensitive enough to detect an integrity breach of 3 micrometers (µm) or smaller. This ensures detection of any breach that could pass a Cryptosporidium oocyst (4 to 6 µm).
2. Sensitivity Criterion: The test must be capable of verifying a Log Removal Value (LRV) equal to or exceeding the pathogen credit awarded by the state primacy agency: $$LRV_{DIT} \ge \text{Removal Credit Claimed (typically 4.0-log)}$$
3. Testing Frequency: Direct integrity testing must be performed on each membrane skid at least once every 24 hours of operation (typically automated during the quietest demand hour of the night).

---

Direct Integrity Testing: The Pressure Decay Test (PDT)

The Pressure Decay Test (PDT) is the standard direct integrity method used in municipal hollow-fiber installations. The test relies on bubble point capillarity: when a hydrophilic membrane's pores are wetted with water, surface tension holds water tightly inside the submicron capillaries. Air applied to one side cannot displace water through intact submicron pores until the applied pressure exceeds the bubble point pressure ($P_{bp}$), defined by the Young-Laplace equation:

$$P_{bp} = \frac{4 \gamma \cos \theta}{d}$$

Where:

• $\gamma$ = Surface tension of water ($72.8\text{ dynes/cm}$ at 20°C)
• $\theta$ = Liquid-membrane contact angle
• $d$ = Pore or defect diameter

For an intact ultrafiltration pore ($d = 0.02\,\mu\text{m}$), the bubble point pressure exceeds 50 to 100 psi. However, for a 3 µm defect, the air bubble point is only 10 to 15 psi. By applying test air at 12 to 25 psi, air cannot penetrate intact pores, but it will immediately bubble through any breach $\ge 3\,\mu\text{m}$.

                        Pressure Decay Test Sequence
                        
 [ Isolate Skid ] ---> [ Drain Permeate ] ---> [ Pressurize Air ] ---> [ Measure Decay ]
    Offline                 Open Vent              15 to 20 psi              5 to 10 min
   Isolate from            Drain Liquid           Fill Fiber Lumen         Decay Rate (psi/min)
   Filtration              from Header            Hold Pressure            Compare to UCL

Step-by-Step PDT Procedure

1. The automated SCADA system isolates the membrane skid from forward filtration.
2. Liquid is drained from one side of the membrane (most commonly the filtrate/permeate header).
3. Oil-free instrument air is introduced to the lumen or feed side until target test pressure ($P_{test}$, typically 15 to 20 psi) is reached.
4. The air supply valve closes, isolating the pressurized volume.
5. The system pauses for a 1- to 2-minute stabilization hold to allow temperature equilibration and fiber expansion.
6. Over a subsequent 5- to 10-minute test period, a high-precision pressure transducer measures the decay in air pressure ($dP/dt$, in $psi/min$).

Upper Control Limit (UCL) and Operator Response

The state primacy agency and membrane manufacturer calculate a site-specific Upper Control Limit (UCL) for each skid, representing the maximum allowable pressure decay rate that mathematically guarantees 4.0-log removal:

• Normal Baseline Decay Rate: Intact skids exhibit a slight pressure decay due to natural diffusive air passage across wetted pores, typically 0.02 to 0.10 psi/min.
• Failed Test: If the measured decay rate exceeds the UCL (commonly 0.25 to 0.40 psi/min depending on skid volume), the skid fails the direct integrity test.

Class II Mandatory Action: A skid that fails DIT must be automatically locked out of service by SCADA. Under no circumstances may an operator manually override the lockout to deliver unverified water to customers. The skid must remain offline until troubleshooting locates and repairs the breach, followed by a passing DIT confirming decay below the UCL.

---

Pinpointing Fiber Breaches and Pinning Repairs

When a skid fails its automated PDT, operators must isolate which specific module houses the breach and pinpoint the individual damaged fiber among thousands of intact strands.

                       Fiber Pinning Repair Sequence
                       
   +--------------------------------------------------------------+
   | 1. Locate Failing Module via Clear Sight Glass or Sonic Tool |
   +--------------------------------------------------------------+
                                  |
                                  v
   +--------------------------------------------------------------+
   | 2. Pressurize Module to 5 - 10 psi; Flood Potting Face       |
   +--------------------------------------------------------------+
                                  |
                                  v
   +--------------------------------------------------------------+
   | 3. Observe Continuous Bubble Stream from Severed Fiber Lumen |
   +--------------------------------------------------------------+
                                  |
                                  v
   +--------------------------------------------------------------+
   | 4. Drive Medical-Grade Titanium/SS Pin into Both Lumen Ends  |
   +--------------------------------------------------------------+
                                  |
                                  v
   +--------------------------------------------------------------+
   | 5. Re-run Pressure Decay Test (PDT) to Confirm Decay < UCL   |
   +--------------------------------------------------------------+

Pinpointing Techniques

1. Acoustic / Ultrasonic Sensor Probing: An operator presses a high-frequency acoustic sensor (sensitive to 35–40 kHz) against the manifold piping or individual module caps while the skid is pressurized with air. Air escaping through a ruptured fiber generates turbulent sonic noise that sounds like a distinct hiss in acoustic headphones.
2. Visual Bubble Testing: The suspect module's top permeate clear end-cap or sight glass is removed, and clean water is poured over the polyurethane potting surface. With 5 to 10 psi air applied to the lumen or feed side, a steady stream of rising bubbles instantly identifies the exact compromised fiber lumen.

Fiber Pinning Protocol

Once a severed or damaged fiber is located:

• Pin Material: Operators must use manufacturer-approved medical-grade titanium or 316L stainless steel tapered pins. Wooden toothpicks, adhesives, silicones, and glues are strictly prohibited because they degrade under chemical CIP and contaminate finished water.
• Dual-End Isolation: The pin is inserted firmly into the open fiber lumen at the top potting face, and a matching pin must be inserted into the corresponding lumen at the bottom potting face. Sealing both ends isolates the severed fiber completely from the clean permeate stream.
• Cumulative Pinning Threshold: Manufacturers establish a maximum allowable pinned fiber limit per module, typically 0.5% to 1.0% of total fibers (e.g., no more than 50 to 100 pinned fibers in a 10,000-fiber module). Exceeding this threshold compromises module hydraulics and demands total module replacement.
• Post-Repair Verification: The module is reassembled, and the operator re-runs the PDT. Only after the decay rate verifies below the UCL can the skid return to active drinking water production.

---

Indirect Integrity Monitoring & Membrane Diagnostics

While direct integrity testing provides a rigorous quantitative check every 24 hours, Indirect Integrity Monitoring (IIM) provides continuous surveillance between daily DIT cycles.

Membrane Autopsy Procedures

When membrane elements suffer chronic fouling, unrecoverable permeability loss, or repeated structural integrity failures, sacrifice modules undergo forensic autopsy:

• Visual Inspection: Disassembling the module or unrolling the spiral element to inspect the feed spacers and active surfaces for discoloration, tearing, and sludge accumulation.
• Dye Penetration Testing: Pumping a dilute dye solution (such as Rhodamine WT or methylene blue) across the membrane under slight pressure. Dye stains reveal pinhole breaches, potting seal debonding, and glue-line delaminations.
• Loss on Ignition (LOI): Foulant cake scraped from the membrane is heated in a muffle furnace to 550°C. The percentage of mass lost represents volatile organic content (biofilms, extracellular proteins, humic acids), while remaining ash represents inorganic scale ($CaCO_3, SiO_2, Fe_2O_3$).
• Scanning Electron Microscopy (SEM) & EDX: High-magnification electron imaging paired with Energy Dispersive X-ray Spectroscopy identifies the exact elemental matrix (e.g., aluminum, silicon, iron, sulfur, calcium) of the foulant layer.
• Fujiwara Chemical Test: Applied to polyamide RO membranes to detect halogen bonding, definitively proving whether irreversible membrane degradation was caused by accidental exposure to free chlorine.