Agenda Item 620-261Date: April 24, 2003

Title: Rules for Full Containment Tanks – Temporary Openings

Handled By:

Doug Miller
Chicago Bridge and Iron Company
1501 North Division Street
Plainfield, IL 60544-8984
Telephone: 815-439-6522
Fax: 815-439-6560
Email: / Richard Whipple
Fluor Daniel
Sugar Land, TX
Telephone: 281-263-3154
Email:

______

Purpose: To permit temporary openings in shell after hydro-test for full containment tanks to Appendices R and Q that have no permanent shell penetrations.

Source: Inquiry 620-I-09/02

Impact: Improved construction safety. Some cost savings as well.

The need:It is becoming increasingly common for low temperature and cryogenic tanks for the storage of fully refrigerated butane, propane, ammonia, liquefied natural gas and similar gases to be stored in “full containment” type tanks. Purchasers specify full containment to maintain control of vapor releases resulting from any primary tank leak. Full containment tanks never have any piping through the shell (all piping is routed through the roof) in order to eliminate the largest risk for leaks or ruptures. Eliminating ALL shell penetrations, including the more commonly used welded manways, is a further requirement on many full containment projects for purchasers who demand a “smooth shell” detail.

During construction, a shell plate is left out creating a "door sheet" opening. This provides ventilation, safe exit routes and efficient access to tank interior. However, once the doorsheet openings are closed on a smooth shell, the lack of openings impedes the safe and efficient execution of the inner tank hydrotest and post-test activities. First, the lines to fill and remove the hydrotest water must be routed up and over the shell. Second, a hazardous condition is created for the workers who dismantle the test lines, remove the in-tank pumps required to pump out the test water, clean the tank, install the in-tank product pumps and foot valves, and complete any other final operations. No shell openings means that the only emergency exit route available to construction workers is through the roof, typically a 100' climb. If there were any sort of incident inside the tank without ground level access openings, it would take hours to rescue a person from the bottom. Further, natural ventilation does not exist. Mechanical forced ventilation can be provided although its quality and reliability is inferior. Since safety is a paramount value, it is worth considering provisions under which temporary openings could be provided after hydrotest on these smooth shell designs.

Justification:It is believed that a smooth shell design containing a restoration plate welded after hydro-static test in accordance with the proposed rules has integrity as high as the same shell containing permanent shell penetrations that have experienced hydrostatic test. The basis of this is because:

1.The proposed restoration welds receive 100% RT and/or UT while attachment welds of shell penetrations have no volumetric NDE so that assurance that any remaining flaws are small is higher for a restored-shell detail. Even more NDE is specified for App R (mandatory UT) in keeping with more stringent h-test requirements and less reserve toughness in carbon steel App R tanks compared to alloy steel App Q tanks.

2.The proposed restored shell has a smooth geometry so that discontinuity stresses are smaller than with shell penetrations.

3.The lower operating temperature and toughness in App R and Q tanks can make the critical flaw size smaller during operation than during the much warmer test conditions. For this reason, hydrotest does not necessarily prove a higher level of assurance against fracture than do improved details and examination.

4.Additionally note that existing Q.8.1 already permits a partial height hydrotest. For an LNG tank this is typically only 60% of full height as controlled in Q.8.1.2 by the foundation (0.48 SG x 1.25 = 0.60). Therefore it is already accepted that the seams in the upper 40% of LNG tank shells receive no hydrotest stress even through they are welded before test.

Proposed Change:Commentary:

Q.8 Testing the Tank in Contact With the Refrigerated Product
The provisions stated in this section are testing requirements for the tank refrigerated by the liquid contents. Provisions noted in Q.9 cover the outer tank, which is not in contact with the refrigerated liquid and is subject to a higher temperature that approaches atmospheric.
Q.8.1 GENERAL PROCEDURE
Q.8.1.1 A thorough check for tightness and structural adequacy is essential for a single-wall tank or for the inner tank of a double-wall tank. Except as permitted by Q.8.6, the test shall be conducted after the entire tank is complete but The hydrostatic test shall be completed before the insulation is applied. Except as limited by foundation or stress conditions, the test shall consist of filling the tank with water to the design liquid level and applying an overload air pressure of 1.25 times the pressure for which the vapor space is designed. Where foundation or stress conditions do not permit a test with water to the design liquid level, the height of water shall be limited as stated in Q.8.1.2 and Q.8.1.3.
Q.8.6 TEMPORARY OPENINGS AFTER HYDROSTATIC TEST
When approved by the purchaser in writing, and only in the case of tanks which when complete have no shell penetrations, it is permitted to restore by welding up to four temporary shell openings after the hydrostatic test in accordance with the provisions of this section.
Q.8.6.1 Temporary openings shall be restored by the insertion of shell plate which matches the thickness and specification of adjacent shell material and is welded into place with full fusion butt welds. The insert plate shall be round with diameter no greater than 42”.
Q.8.6.2 The insert plate weld shall not cross any shell seams and shall be at least the greater of 10 times the shell thickness or 12 in. from any other weld in the shell including shell seams, shell-to-bottom weld or attachment welds.
Q.8.6.3 The butt weld around the periphery of the plate shall be examined over 100% of its length by both liquid penetrant method and radiographic method. The liquid penetrant examination is required on the root pass, on the backgouged surface, and on the inside and outside finished weld surfaces. Additionally, the weld shall be vacuum box leak tested.

620-261page 1 of 4