How Is a Leak Between Walls Detected: Interspace and Spark Test Evidence
A leak between the walls of a double-wall tank is detected by monitoring the interspace, not by hoping it stays dry: install a level or vapour sensor in the annulus, set an alarm and a trip at a fraction of the interspace width, and test the annulus at commissioning so a dry reading means monitored rather than untested. On a single-wall shell the same event is caught by a drop in inventory that the flow balance cannot explain, backed by a 1500 V DC spark test of the enamel panels and a holiday survey of any coated surface. Detection is a monitoring design decision taken before the tank is ordered.
A double-wall storage tank exists so that a leak is caught before the product reaches the ground. That only works if the space between the walls is actually monitored: a sensor, an alarm and someone who knows what the reading means. Most double-wall installations fail on that last part — the annulus is dry on the inspection date and everyone concludes the tank is sound, when in fact the monitoring system was never commissioned or the alarm was never connected. The engineering question is therefore not "how sensitive is the sensor" but "what proves the detection path works". Answer that at design and the rest follows, because the real question how is a leak between walls detected is answered by the monitoring design rather than by the sensor rating.
The Detection Path Has Three Stages
Commissioning, continuous monitoring and periodic verification. At commissioning the interspace is proof-tested and its dry condition is recorded, so that a later wet reading is meaningful rather than ambiguous. In service the interspace is monitored by level, by conductivity, or by a vapour sensor where the product is volatile; the alarm is set well below the full interspace width so there is time to respond, and the trip closes the inlet. Periodically the whole path is verified: the sensor is tested with a known signal, the alarm path is checked, the drain and the sump are cleared, and the record is filed. Skipping the first and third stages is what turns a double-wall tank into a single-wall tank with a second wall.
On a single-wall shell the same event manifests as an unexplained inventory loss. The flow balance — received minus pumped minus discharge — leaves a figure that no meter accounts for, and that figure is the signal. The practical check is a level drop test with the tank isolated: close the inlet and the outlet, record the level against a knowngauge over a defined period, and confirm it is stable within the instrument tolerance. On a bolted shell, a weeping joint shows as a damp track down the panel or a fresh stain at the gasket line, and on an enamel shell a pinhole shows as a local weep rather than a spreading blister because the enamel is fused rather than attached.
Factory Testing Already Rules a Fraction of the Risk
The panel-level test eliminates the largest class of leak before erection. Every enamel panel is spark tested at 1500 V DC before shipment, which finds the holiday that would otherwise become a pinhole weep in service; thickness is recorded per panel; and the joint is sealed with 8.8-grade bolts and an EPDM gasket at a specified torque, recorded at assembly. A leak at commissioning therefore points at the joint, the foundation or the nozzle rather than the enamel — and the joint is the most inspectable part of the assembly. That is why a torque record, not a visual check, is the document that matters when a weep appears at first fill.
Technical Specification
Parameter | Typical Value / Range | Note |
Enamel holiday detection | 1500 V DC spark test | per panel before shipment |
Enamel layer thickness | 0.25–0.45 mm | rated above 3450 N/cm² |
Joint | 8.8-grade bolts, EPDM gasket | torque recorded at assembly |
Interspace alarm set point | a fraction of interspace width | gives response time before filling |
Level drop test | isolated tank, defined period | confirms no unexplained inventory loss |
Coated shell survey | holiday detection on the film | for epoxy to AWWA C550 at 180–280 µm |
Design life | ≥ 30 years | with the specified inspection regime |
Inspection interval | visual at each shutdown, torque after first year | documented in the operation manual |
Designing the Interspace So It Can Be Monitored
The annulus has to be reachable, drainable and instrumentable. Provide a drain or a sump at the low point with a clear path to atmosphere, so the space can be emptied for inspection and so that liquid accumulating in a low corner is detected rather than simply sitting there. Give the interspace a vapour monitor where the stored product has a vapour, and a liquid level or conductivity probe where it does not; fit the taps and the penetrations in the inner wall where they can be isolated and tested; and keep the annulus free of materials that hold water against the outer wall. Where the inner shell is a bolted enamel tank on a ring beam, the space under the floor and around the ring beam is part of the same question and needs drainage too, because trapped water there is a corrosion source even when the product side is dry.
What the Operator Checks and Records
A short, fixed routine converts detection into evidence. At each fill verify the interspace reading against the commissioning record; at each shutdown inspect the shell interior with a torch, checking particularly the gasket lines, the panel edges and the nozzle penetrations; check the torque record and the bolt heads at the first year and again at each major shutdown; survey any coated shell surface with a holiday detector on the cadence the specification sets; and keep a weep-and-repair log with the date, the location and the corrective action. Thirty years is the design life, and this routine is what makes a leak found early a two-hour repair instead of a shutdown.
Project Case
Project | Location | Product | Capacity | Scope |
Big-diameter water tank | Namibia | GFS tank | 44,900 m³ | supply + installation supervision |
Potable water series | Indonesia | GFS tank | 21,099 m³ (φ42.04 m × 15.2 m) | supply + commissioning support |
Firewater series | Sichuan, China | GFS tank | 8,930 m³ (φ19.87 m × 14.4 m × 2) | supply + installation supervision |
Industrial wastewater series | Xinjiang, China | GFS tank | 30,469 m³ in 27 tanks | supply + installation supervision |
Center Enamel Engineering Capability
Center Enamel (Shijiazhuang Zhengzhong Technology Co., Ltd) has designed and fabricated bolted storage tanks since 2008. As the first glass-fused-to-steel (GFS) tank manufacturer in China, the company holds close to 200 enamel-related patents, produces roughly 300,000 enamel-coated steel plates a year, has completed more than 30,000 installed projects and supplies its tanks to over 100 countries. The new 150,000 m² production base was added to raise output capacity, and single tanks are supplied up to 60,000 m³. Manufacturing runs under ISO 9001 and ISO 45001, with product certification including NSF/ANSI 61, WRAS, FDA, LFGB, CE (EN 1090), ISO 28765, FM, BSCI and EUROCODE, and design referenced to AWWA D103-09, AWWA C550 and NFPA where the application requires it.
For leak detection the capability lands as panel-level 1500 V DC spark test records that remove pinholes before erection, bolted joints delivered with torque records, and interspace taps, drainage and instrumentation arranged so the monitoring path can be commissioned and verified rather than assumed.
Frequently Asked Questions
Q1: What is the alarm set point in the interspace?
A1: A small fraction of the interspace width, so there is measurable time to isolate the tank and investigate before the space fills. The exact value should be stated in the specification and justified by the response time the site can actually achieve.
Q2: How is a dry interspace reading at commissioning verified?
A2: By proof testing the interspace and recording its dry condition with the date, so the baseline is documented. Without that baseline a later reading cannot be compared with anything.
Q3: How is a leak found on a single-wall shell?
A3: By the inventory balance. Incoming minus outgoing minus pumped leaves a figure no meter explains; confirm it with an isolated level drop test over a defined period, then inspect the gasket lines, the panel edges and the nozzle penetrations.
Q4: Does the enamel panel test catch the leak risk?
A4: Yes for pinholes. The 1500 V DC spark test on every panel before shipment is designed to find exactly that defect, so a weep at first fill usually points to the joint or a foundation settlement rather than the enamel itself.
Q5: Can a bolted shell be fitted into a double-wall arrangement?
A5: Yes, but the detail has to be designed including the ring beam drainage and the space around the foundation, because water trapped under and around the shell is a corrosion source even when the product side is intact.
Q6: What is surveyed on a coated shell?
A6: A holiday detection survey of the film, with the count and location logged. The same principle as the enamel spark test, applied to an epoxy coating at 180–280 µm to AWWA C550.
Q7: How often is the detection path tested?
A7: At commissioning with a known signal, then at each major shutdown test the sensor and the alarm path, verify the interspace drain is clear, and re-check the torque record; the weep-and-repair log closes the loop.
Leak detection is a monitoring design, not a sensor spec: a reachable and drainable interspace, an alarm set with response time in mind, a commissioning baseline, and a periodic verification of the whole path. On the product side, the 1500 V DC spark test on every enamel panel and the torque record at assembly remove the two most likely leak sources before the tank is filled, and an unexplained inventory loss plus an isolated level drop test finds the rest. Write the detection scope into the tank order and the double wall does its job; leave it out and the second wall is decorative. Send the product, the wall arrangement and the site requirements, and the detection design follows.
Talk to an Engineer
Send the product and its vapour behaviour, the single-wall or double-wall arrangement, the interspace width and access, the available instrumentation and alarm routing, the tank diameter and height and foundation type, and the environmental release limits. The engineering team will return a leak detection scope with the monitoring method and alarm set point, the commissioning test procedure, the interspace drainage and tap layout, the panel test records to expect, and the inspection routine. Where a different shell arrangement fits better, that option is stated alongside rather than replacing it.