API 598 shell and seat tests are the two acceptance points that determine whether a high pressure swing check valve order ships as agreed. If the purchase order only says "tested per API 598," the supplier still has to interpret what to pressure-test, what counts as a pass, and which records arrive with the shipment. A vague clause may not stop the valves from being sound, but it leaves acceptance without a signed record to check against. Understanding what shell and seat tests cover, how the test pressures line up on a high pressure swing check valve, and how to word the requirement so the paperwork matches the delivered hardware makes the order enforceable.
Why Backflow Risk Makes Pressure Testing a Commercial Requirement
Swing check valves sit in pump discharge lines, steam headers, and process piping where flow should move in one direction. When a pump trips or header pressure drops, the disc swings shut against returning flow. If that closure is late or the seat does not hold, reverse flow accelerates back through the line. Steam systems show the mechanism clearly: returning condensate slams into the pipework and produces the pressure spike known as water hammer. The same effect appears in liquid service as a surge that loads flanges, supports, and the pump from the wrong side. The cost shows up after the event. A leaking valve can mean an unplanned shutdown, a cut-out spool, a crane lift, and a re-test of the reinstalled line. The valve itself is rarely the expensive part; the outage is. Pressure testing belongs in the order because it creates a documented check before installation. National Board guidance on high pressure fluid valve integrity management follows the same logic: a documented test before installation costs far less than finding a leak in service, and it gives the buyer something concrete to review when the shipment arrives.
Understanding Shell Tests and Seat Tests Under API 598
API 598 divides the work into two test families. The shell test pushes pressure into the body cavity with the seat isolated, so it grades the pressure boundary. The seat test, often called the closure test, checks whether the disc and seat hold tight—first at high pressure and then at low pressure. Both tests run on the assembled valve at the factory before painting and packing.
1. How Shell Testing Protects the Valve Body Before Delivery
The shell test pressurizes the assembled body and bonnet with the disc held clear of the seat, so the full pressure boundary is loaded: casting walls, the body-to-bonnet joint, bolting, and any penetration. Test pressure runs above the nominal rating. The published table for these valves gives a useful reference: a PN16.0 valve is shell tested at 24.0 MPa, with smaller ratings scaling down to PN1.6 at 2.4 MPa. Water is the usual medium because it is safe to handle, visible when it leaks, and easy to hold at pressure. The shell test catches what a visual check cannot: casting porosity, a marginal gasket squeeze at the bonnet joint, and bolting that relaxes under load. A signed shell test record makes the pressure boundary verifiable at the factory gate rather than at the flange in the field.
2. How High and Low Pressure Seat Tests Support Buyer Acceptance
The closure test speaks to the check valve's actual job. In the high pressure closure test, water at roughly 1.1 times the nominal rating is applied against the closed disc—17.6 MPa for the same PN16.0 valve—and leakage past the seat is measured. The low pressure closure test repeats the check at 0.6 MPa with air or gas, which is where small seat leaks show up that a water column can mask. API 598 sets an allowable leakage rate for each valve type and size, and metal-seated industrial check valves are judged against that rate. On the acceptance side, traceability matters: valve serial number, test medium, test pressure, hold time, measured leakage, and the inspector's signature on the same record.
Writing API 598 Test Requirements into an RFQ or Purchase Order
A test clause works when it names the standard, the tests, and the acceptance basis together. Write "shell test and high and low pressure closure test per API 598" rather than "pressure tested," then add the details the standard leaves to the order: test medium, whether your inspector or a third party witnesses the test, and which documents ship with the valves. API 598 is a test standard, not an official product certificate. The evidence you collect is the test record itself, together with the material certificates and inspection documents agreed in the order. A check valve supplier that can share the report format with the quotation is easier to qualify. For wholesale check valve schedules, the same clause should follow every unit. A check valve manufacturer should be able to state the test medium, pressure, hold time, and acceptance basis without ambiguity. For wholesale swing check valve orders, the clause has to scale across sizes and pressure classes. A swing check valve manufacturer that controls this documentation can make factory acceptance a comparison against agreed terms instead of a debate about what "tested" was supposed to mean. Witness testing and non-standard documentation take time, so build the review window into the schedule from the start. These valves are produced against order, with a reference production period of 15–30 days. A witness plan that arrives two days before shipment is a plan that holds the container. State the size, pressure rating, end connection (flanged to ASME B16.5, ASME B16.47, EN1092, or GB/T 9113; buttwelding to GB/T 12224), body material (WCB, CF8/304, WC6, or LCB), and design standard (API 6D, ASME B16.34, BS1868, GB/T 12236) in the same document, so the tested build and the delivered build are identical. For the high pressure swing check valves we supply, testing is executed to API 598 alongside API 6D, JB/T 9092, and GB/T 13927, with pressure coverage from PN1.6 up to 42.0 MPa and service temperatures from -46°C to 570°C. Send the RFQ with pressure class, size, medium, end connection, material, and the test documents you need, and ask for the shell and closure test report format with the quotation. That request gives you a practical view of a supplier's process control.
Conclusion
Pressure testing earns its place in the purchase order because it turns a product claim into a signed record for the exact valve that shipped. Shell testing grades the pressure boundary; high and low pressure closure testing grades the seat that stops backflow. Both are judged against API 598 leakage allowances rather than an ideal of zero. Agree the tests, the documents, and any witness points early, and factory acceptance becomes a check against agreed terms instead of a debate about what "tested" was supposed to mean.
FAQ
Q：What does API 598 testing cover for a swing check valve order?
A：It covers two things: a shell test that pressurizes the body and bonnet with the seat isolated to grade the pressure boundary, and a closure or seat test run at both high pressure (about 1.1 times the nominal rating) and low pressure (0.6 MPa air or gas) to measure leakage past the closed disc. Both run on the assembled valve before shipment, and the results are recorded for each unit.
Q：Should I require API 598 shell and seat test records with my purchase order?
A：Yes, when acceptance matters. A general "pressure tested" line leaves nothing to compare the delivered valve against, while a record showing serial number, test medium, test pressure, hold time, and measured leakage lets a factory acceptance manager confirm the valve matches the order. Put the records in the purchase order as an agreed delivery document so the expectation is clear from the start.
Q：How should API 598 test requirements be written in an RFQ?
A：Name the standard and the tests directly—"shell test and high and low pressure closure test per API 598"—then add the order-specific details: test medium, whether your inspector or a third party witnesses the test, the acceptance basis, and the document package expected with shipment. Include size, pressure class, end connection, and body material in the same section so the tested build matches the delivered build.
Sources / References
Water Hammer: The Mechanism | TLV
Technical Articles | National Board
Software Quality Assurance Program and Guidelines (NUREG/BR-0167) | Nuclear Regulatory Commission
No comments:
Post a Comment