THE TRAGIC LOSS OF THE NUCLEAR SUBMARINE THRESHER 10 APRIL 1963
CAPT ZACK T. PATE, USN, RET.
RADM DAVID GOEBEL, USN, RET.
VADM GEORGE EMERY, USN, RET.
Released: June 2018
The USS Thresher SSN-593 Underway 1961 (Lost 55 years ago - 10 April 1963)
The USS Thresher (SSN-593) was the first in a new class of deep diving fast attack nuclear submarines. She was constructed in the Portsmouth Naval Shipyard in Kittery, Maine. By early April 1963 Thresher had completed her “Post Shakedown Availability” (PSA) and went to sea for PSA sea trials. As part of those sea trials a dive to test depth was scheduled. Thresher did not recover from the deep dive and sank to the ocean floor at a depth of just over 8000 feet. All 129 persons on board died instantly when the ship’s hull collapsed/imploded at a depth estimated to be at least 150% of test depth. Extensive investigations were conducted in the following weeks and months, including by a Navy Board of Inquiry, which was promptly established at the shipyard, and later extensive congressional hearings were held. The only other US nuclear submarine that has been lost at sea is the USS Scorpion (SSN-589), which sank in the mid-Atlantic on 22 May 1968. While there has been much research and speculation, the cause of the loss of Scorpion is still a mystery. A battery explosion is often cited as a factor, but the cause of the battery explosion is unknown. The cause of the loss of Thresher is much better understood for two principal reasons: Thresher had an escort ship and a SOSUS (off shore Sound Surveillance System) array was located in relatively close proximity. In Thresher’s case that array was on the ocean floor south of Nova Scotia, and about 30 nautical miles from the accident site. Virtually all information about the tragedy was quickly classified, including the report by the Board of Inquiry, congressional hearings, SOSUS data, as well as the underwater telephone (UQC) communications with the escort ship, USS Skylark (ASR20). This information was (page-break)
THE SUBMARINE REVIEW JUNE 2018 126
de-classified (slowly and piecemeal) over time and now only minor details remain classified. Much has been written about the tragedy, especially since the de-classification of the records began. In the view of the three co-authors of this article, there is of yet no short, easy to comprehend account that is both as factual as possible and unbiased. The three co-authors have studied and discussed the available history extensively and have sought to achieve just such an account. Here is our analysis of Thresher’s deep dive and loss. ************
Thresher’s final 25 minutes:
Thresher reports her depth (test depth -300 feet), going to test depth, to her escort ship, USS Skylark (ASR-20), on the Under Water Telephone (UQC).
Skylark reports to Thresher on the UQC that she is changing course to 090 degrees. At about this time, plus or minus a few minutes, Thresher reaches test depth.
0909 The Sound Surveillance System (SOSUS) detects what analysts describe as “line frequency instability”. This most likely means that Thresher’s Ships Service Turbine Generators (SSTGs) were not stable at 60 Hz, as they normally are.
0909.8 SOSUS detects what some analysts describe as a main ballast tank (MBT) blow of 1.5 minutes duration.
0910 Thresher reports a course change in what appears to be a routine UQC communication. Since the report was made without alarm the classification of the noise detected by SOSUS at
as a MBT blow is questionable. Additionally, Skylark did not report detecting MBT blow noise at this point, but both SOSUS and Skylark did detect and reported MBT blow noise three minutes later.
THE SUBMARINE REVIEW JUNE 2018 - 127
As has been described in other sources (Note 1), attempts by Thresher to blow main ballast tanks were so ineffective that VERY little water was being displaced from the ballast tanks, and the air flow through the ship’s airlines was only a small fraction of designed air flow. Thus, the noise transmitted to sea could be very different than the noises detected by SOSUS in other MBT blow situations.
SOSUS detected Thresher’s Main Coolant Pumps (MCPs) stopping. (SOSUS had detected MCPs running in fast speed earlier in the dive). The loss of MCPs in fast speed could be due to instabilities in SSTG output, as noted by SOSUS two or so minutes earlier, or due to a reactor scram. In any event, a reactor scram almost certainly occurred at this point. The abrupt stoppage of Main Coolant Pumps (without the normal procedure of first shifting the pumps to slow speed) would cause an immediate reactor scram as the reactor protection circuitry detected a complete loss of flow. The reactor scram could also have been caused by salt water spray from leakage from any one of the MANY sea water valves / fittings / flanges / sil-braze joints / valve stem packings / etc. located in the machinery space. The machinery space on Thresher was located between the reactor compartment and the engine room and contained extensive sea water system piping. Reactor control electronics and the AC/DC motor generators were located in this space. The ship has now lost all propulsion. Shifting propulsion to the ship’s Emergency Propulsion Motor (EPM), a low powered electric motor wrapped around the shaft to the propeller, was a slow process. Shifting to the EPM required dis-engaging a massive clutch in the shaft so that the EPM would be turning just the propeller and not the main turbine rotors and the reduction gears. On Thresher this had to be done locally, at the very stern of the ship. And, at that point in history, Naval Reactors procedures did not make provisions for use of residual heat in the reactor for propulsion.
Thresher responded to a UQC transmission check from Skylark without expressing alarm. Both ships used the code word “KILO” meaning that communications were satisfactory. At that point, however, (page-break)
THE SUBMARINE REVIEW JUNE 2018 128
the Captain and Officer of the Deck, both in the Control Room, would very likely have been getting alarming and, at least at first, confusing reports from the Maneuvering Room, located aft in the engine room.
Thresher reports on the UQC: “Experiencing minor difficulties. Have positive up angle. Attempting to blow up.” This transmission is thought to have been made by the Commanding Officer. The ship was using coast down speed after the scram to plane up and was likely able to get at least a few feet above test depth. The ship was heavy from sea water leaks and possibly from being out of trim---not noticeable at 10 or 12 knots, but very noticeable at 2 or 3 knots.
Skylark and SOSUS both “…detected noise thought…” to be from Thresher’s attempt to blow Main Ballast Tanks. The noise had a duration of 30 seconds. With no propulsion and an ineffective MBT Blow System -- and heavy -- the ship was slowly sinking,
Thresher reported “exceeding test depth” on the UQC. (In this transmission the words “test depth” were clear while the first word was garbled; but widely thought to be “exceeding” or “passing”.)
Skylark and SOSUS heard the hull implode, instantly killing 129 souls. ----end---
Notes: (1) In a letter and statement of analysis to the president of the Submarine Museum and Library, written on the 50th anniversary of the loss of Thresher, retired Captain Zack T. Pate describes experience on Thresher’s sister ship at the shipyard, the USS Tinosa (SSN-606), which provides data and physical evidence that convincingly demonstrate that Thresher’s Main Ballast Tank blow system was seriously degraded. The principal cause of the degradation was “temporary” strainers in the air system that collapsed. Fine mesh conical strainers, backed by a thick bronze plate with a small orifice, were installed by the shipyard to protect air system valves from minute particles of debris during construction. The strainers were not removed before the ship went to sea, as intended, (page-break)
THE SUBMARINE REVIEW JUNE 2018 129
and the conically shaped mesh collapsed into the back-up orifice, blocking air flow to the main ballast tanks. See the photo below. The aforementioned letter and statement are included in this issue of The Submarine Review and are also available at the Submarine Library and Museum Association in Connecticut. (2) The authors do not believe that Thresher suffered a major flooding casualty, as suggested by the Navy Board of Inquiry. SOSUS experts believe that the noise from major flooding at test depth would have been unmistakable at the (fairly nearby) SOSUS array and, in addition, the report by the Captain of “experiencing minor difficulties” is inconsistent with major flooding. No submariner considers flooding a minor event. We do believe, however, that it is likely that Thresher experienced sea water leakage and, that even a relatively small leak at test depth could cause considerably spraying of salt water onto electrical and electronic equipment. We consider that a probable cause of the frequency instability detected by SOSUS and, directly or indirectly, of the reactor scram. (3) In this article the authors have not sought to assign responsibility / accountability. We believe, however, that there is plenty to go around, (page-break)
THE SUBMARINE REVIEW JUNE 2018 130
and that the shipyard is certainly not exempt. All three co-authors served in the Navy’s nuclear submarine program as the principal part of their career. VADM Emery served as Commander Submarine Force Atlantic before retiring. RADM Goebel served as Commander Submarine Group Two and now, in retirement, serves as President of the Submarine Library and Museum Association in Connecticut. CAPT Pate served as a special assistant to Admiral Rickover before retiring. CAPT Pate’s first assignment on a nuclear submarine was as Auxiliary Division Officer on USS Tinosa (SSN-606), the sister ship to Thresher at the Portsmouth Naval Shipyard. After Thresher’s loss then LT Pate was given a special assignment by his commanding officer to serve as liaison to the Navy Board of Inquiry and with shipyard engineers in the search for the cause of Thresher’s loss; AND in the intensive quest to make sure Tinosa did not suffer the same fate. Tinosa did extensive tests alongside her pier at the request of the Board of Inquiry, especially of the Main Ballast Tank Blow System. (see Note  above).CAPTAIN Zack Pate’s letter to RADM Goebel on the subject of the Main Ballast Tank blow system investigation following the loss of the USS Thresher May 31, 2013 RADM David Goebel 771 Pequot Avenue New London, CT 06320Dear Dave,The purpose of this letter is to forward a strainer (and its associated orifice) which was removed from the 4500/3000 psi air system on the USS Tinosa (SSN-606), shortly after the loss of USS Thresher (SSN593) in 1963. Tinosa was Thresher’s sister ship, under construction at the Portsmouth Naval Shipyard, when Thresher was lost at sea on 10 April 1963. The Thresher was the first ship of a new class. The Court of Inquiry, formed to investigate the loss of Thresher, (page-break)
THE SUBMARINE REVIEW JUNE 2018 131
asked Tinosa to conduct tests of the main ballast tank blow system shortly after the court was formed. At the time I was the Auxiliary Division Officer (aka Damage Control Assistant, or DCA) on Tinosa. When the main ballast tank blow system failed to operate as designed, auxiliarymen working for me soon located collapsed strainers just upstream of both 4500 to 3000 psi (Marotta) reducing valves. The collapsed strainers were quite clearly blocking most air flow to the main ballast tanks, as is evident from a visual inspection of the enclosed strainer, one of the two removed after the test. The two strainers were virtually identical in appearance and condition. After the (failed) air system tests on Tinosa, I was interviewed by CAPT Osborne, a member of the Court of Inquiry. During our discussions I showed him the two collapsed strainers and where they came from in Tinosa’s systems. At the end of the interview I gave one of the strainers to CAPT Osborne to be part of the court’s evidence. With the approval of my commanding officer I kept the other one “until further notice.” Now, 50 years later, and near the 50th anniversary of the tragic loss of Thresher, our sister ship, I am pleased to donate this second strainer to the Submarine Force Library and Museum Association. Immediately following my signature on this letter is my “testimonial”, as you requested, describing key aspects of the special tests on Tinosa, which I personally supervised, and other directly related observations. The strainers were supposed to be a temporary installation, intended to protect the Marotta reducers during construction. There is abundant evidence that the strainers found on Tinosa were still installed on Thresher on the fateful day of her loss. In my view the strainers were a significant, if not critical, factor in Thresher’s inability to recover from problems encountered during her deep dive. After reading the testimony below, I believe most readers will agree. Warmest Regards, (orig. signed) Zack T. Pate CAPT. USN Retired (page-break)
THE SUBMARINE REVIEWJUNE 2018 132
Observations related to the loss of USS Thresher (SSN 693) on 10 April 1963 One of the few communications (by underwater telephone) from Thresher to the accompanying Submarine Rescue Vessel (ASR), during her fatal deep dive, was “experiencing minor difficulties, attempting to blow”, thought to be transmitted by the Thresher’s commanding officer. In addition to this evidence of a problem, the SOSUS recordings during Thresher’s deep dive showed a pattern of intermittent noise that suggested interruptions of air flow. Based on this information (and perhaps other factors), the Court of Inquiry asked Tinosa to conduct tests of its main ballast tank (MBT) blow system.
Thresher was the first submarine to have 4500 psi air banks (cylinders located in the MBTs). The previous maximum was 3000 psi and prior submarines had 3000 psi air systems running throughout the ship, without a need for reducing valves. All of Thresher’s (and Tinosa’s) air systems were also 3000 psi; but two large 4500/3000 psi reducers served the 3000 psi systems from the 4500 psi storage cylinders. Thus, the 3000 psi air used to blow the MBTs when surfacing went through these 4500/3000 psi reducers. When Tinosa was preparing to fully test the MBT blow system alongside a pier (Tinosa was not yet ready for sea, but the air systems had been completed), a team of Shipyard engineers was assigned to help us with instrumentation, recording, etc. Through this team, the Shipyard expressed concern that a full pressure blow of MBTs alongside a pier (rather than submerged) might bulge or even rupture the MBTs. For this reason, Tinosa’s first test of the MBT blow system was done with only 500 psi in the 4500 psi air banks; and instrumentation was placed on a MBT vent cover, topside on the ship, one forward and one aft. [Note that this concern by the Shipyard is compelling evidence that a full test of the MBT blow system was not scheduled or conducted on Thresher before she went to sea. And, in fact, no such test was scheduled or planned for Tinosa at that point in time.] When the test was conducted on Tinosa with 500 psi in the air banks, (page-break)
THE SUBMARINE REVIEW JUNE 2018 133
there was essentially no measurable back pressure in the MBTs during the test. The test was once again conducted with 1000 psi in the air banks, and once again the back pressure in the MBTs was nominal; about 1 psi as I recall. Tinosa was then authorized by the Shipyard to do the full pressure test requested by the Court of Inquiry; and we charged all air banks to 4500 psi. With the air systems instrumented, including timers, we then conducted the full blow test simulating what a submarine would do at sea in an emergency. When the MBT blow valves were opened on order there was a huge rush of air noise in the control room, then silence after a few seconds. Than a few seconds later another brief rush, then silence again. This pattern was repeated over about a six-minute period until the blow valves were closed and the test terminated. Air pressure in the air banks had only dropped a few hundred pounds; to about 4100 psi as I recall. We were shocked. The engineers and Tinosa personnel expected the air banks to be depleted in two or three minutes. We had witnessed less than one-tenth of the simulated deballasting (removal of sea water from the MBTs) that we expected. An auxiliaryman who was stationed in the space where the two 4500/3000 psi reducers were located quickly reported that the reducers and upstream piping were covered in ice. The cause of the icing and reduced air flow was soon found --- a strainer installed upstream of each reducer had collapsed, effectively blocking over 90% of the expected air flow. No Tinosa crew members, officer or enlisted, were aware that the strainers were installed --- nor were any of the Shipyard personnel involved in the testing. We were subsequently advised by the Shipyard that the strainers were installed at the request of the Marotta Valve Company because construction debris was interfering with the ability of the reducers to completely shut off air flow, thus tending to over-pressurize the 3000 psi air system. The strainers were supposed to be removed before the ships went to sea. (page-break)
THE SUBMARINE REVIEWJUNE 2018 134
The strainers were conically shaped (before collapsing), pointed into the airstream, and were backed by a thick brass or bronze orifice with a small center hole (about ¼ inch diameter). The purpose of the orifice was to reduce the pressure differential across the strainer, so it would not collapse. This worked during extensive routine use of the 3000 psi air system on Tinosa and during the MBT blow tests with 500 psi and 1000 psi in the air banks. But when the full pressure MBT blow test was conducted the strainers collapsed. Thus, this observer’s impression is that the strainers on Thresher may have collapsed when an emergency blow of MBTs was commenced or may have been already collapsed or partially collapsed as the collapsed strainers would pass enough air for routine 3000 psi air service and for a brief burst of air when surfacing from periscope depth when the Low Pressure Blower was the primary means of deballasting. So, it is quite possible that Thresher had been operating with collapsed or partially collapsed strainers for some time. There was abundant evidence that the strainers were installed on Thresher, including: • SOSUS patterns that showed the intermittent bursts of air noise as ice formed in the collapsed strainer and then blew out, then reformed, etc.; closely matching the pattern seen during tests on Tinosa. • Testimony by Shipyard personnel to the Court of Inquiry that the strainers were installed on the Thresher. • Thresher Captain Wes Harvey’s statement “attempting to blow” on the underwater telephone • The presence of the strainers on Tinosa In assessing the impact of the degraded MBT blow system on Thresher’s attempts to reach the surface, it is important to understand the role of the propulsion system. Normally, a “Full Bell” (quickly resulting in high speed of the ships propeller) would result in a rapid drive toward the surface. The data available strongly suggests that the reactor scrammed early during Thresher’s plight, and high levels of thrust from the propeller were not available. Subsequent to the loss of Thresher, Tinosa conducted tests (as did other ships) showing that use of decay heat from the reactor could provide a moderate level of propulsion for a THE SUBMARINE REVIEWJUNE 2018135 few minutes after a scram – enough propulsion to be an important asset in an emergency such as Thresher experienced. But this capability was not available to Thresher. Naval Reactor’s procedures did not permit use of decay heat at the time. (Soon after the Thresher loss, the procedures were changed.) Indeed, on Tinosa, and I expect on Thresher, the scram procedure required the upper level engine room watch to trip the main turbine inlet valves locally--- to be re-set only after a full recovery from the scram, at least 20 or 30 minutes later. The Captain of Thresher most likely did not have the option of overriding procedures and using propulsion anyway. Thus, the Main Ballast Tank blow system was the only means of bringing the ship to the surface, and the evidence is strong that it was seriously degraded. When one inspects the enclosed strainer, disabled may seem more accurate than degraded.NOTES: 1) During the tests of Tinosa’s MBT blow system, the question arose as to whether the results of a full blow test on the surface at dockside would closely simulate a full blow at or near test depth. The shipyard (and BUSHIPS at the time) determined that flow to the main ballast tanks would be limited by “sonic choking” in the relatively small 3000 psi air lines to the MBTs; not by back pressure in the MBTs. 2) Three colleagues, who were shipmates on Tinosa during the period under discussion, reviewed this letter and the above observations and assisted in the quest for accuracy. These were:
a. Clarence E. (Ed) Moore, CAPT. USN retired. Ed was Chief Engineer on Tinosa
b. William D. (Bill) Smith, ADM. USN retired. Bill was Electrical and Reactor Control Officer on Tinosa
c. Walter J. (Walt) Coakley, CAPT. USN retired. Walt was Weapons Officer and an Engineering Officer of the Watch on Tinosa
NOTE: To post this information Webmaster took the liberty to note page breaks in the original document and to utilize the word 'cont.' where necessary to note a continuance after the page-break where time-marks made it necessary. Original PDF (in its entirety) available upon request.