ATLAS

579th STRATEGIC MISSILE SQUADRON

WALKER AFB

SITE 5 ACCIDENT FEBRUARY 13, 1964

January 12, 1999

I am writing this account of my recollections for the purpose of providing this information to anyone who wishes to use it. I recognize that it may not be readily understood by those who are not already initiated into the weapon system, but who else would be interested, anyway. I would be happy to communicate with anyone who has interest in or would like more information about the incident which I describe herein. My address is:

Allan L. Kane

6609 S.E. 160th St.

Olalla, WA 98359

AKane42533@aol.com

With the draft breathing down my neck, I enlisted in the Air Force on Dec. 14, 1961 and, following Basic Training, was assigned to Shepard AFB and attended the missile training school there. Following this I went to Walker AFB, arriving in Oct. 1962. I was assigned to Crew #60 and we traveled to VANDENBERG AFB in April, 1963 for our operational training. Shortly after our return to Walker my crew was deemed to be adequately qualified and we began regular duty rotation at the various complexes.

I have very little in the way of documentation and the majority of what I write is based upon recollections alone, and not hard data or documents. The only things I have are my note made just prior to leaving the Launch Control Center following the incident which destroyed the silo. On this scrap of paper I wrote the number 61-2475 which no longer has any meaning to me. It may identify the installation and I must have considered it to be something of significance to that particular site. I also wrote SM-65F, which is the military designation for the Atlas “F” weapon system, 13 Feb. 64, which is the date of the incident, 11:10:22, which is the time of day at which the clock in the LCC stopped, R-60, which was our crew number and Capt. V.P., Ford, Jr., the Crew Commander. I also have some artifacts from the site which I picked up as I left the site that day and which I collected during subsequent trips to the site. These artifacts consist of pieces of the missile skin, hydraulic system components (fittings, valve handles, component ID tags), an elapsed time clock powered by 400 cycle source (probably from a guidance system component or the 400CPS generator), the launcher platform ID plate, an ID plate from a Worthington compressor and a piece of steel decking plate (probably part of the launcher platform).

It was generally known, prior to the actual date of the test, that there would be a PLX at our site in the near future, but neither I nor other crew member knew when it would occur. It was clear that our site had been chosen, as the operational warhead had been replaced with a dummy unit, the engine ignitors had been replaced with the fuse block and other typical preparatory activities had been accomplished during the days preceding the test. As I recall, we were notified at our pre-departure briefing of the intent to initiate the PLX during our tour, but I am not certain of this. It was clear, however, as soon as we arrived at the site that the PLX was scheduled for our tour. There were a number of extra people at the site, or they arrived shortly after we did, who were obviously there to observe the event. I recall that our Sector Commander was there, along with others of the rank of Colonel and below. As I recall, we relieved Major Goon’s crew that day and I recall some of his crew members expressing relief that they had not drawn the short straw.

Our crew composition on that day was as follows:

MCCC - Capt. Vincent Paul Ford, Jr.

DMCCC - 1LT Howard Jones

BMAT - SSGT William Jenkins

MFT - A2C Allan L. Kane

EPPT - A2C (or A1C) Wayne Egnew

I remember being in the vicinity of the LCC prior to the initiation of the countdown, with all crew members present. Capt. Ford and Lt. Jones were discussing the procedure and how they felt about it. Ford offered that he was “as nervous as a f---ing cat” He remembered shortly after having said this that all conversations were being recorded and was sincerely mortified. He generally tended to be of a nervous temperament, and he clearly was now. This is not intended as a criticism, as we were all pretty apprehensive about what we were about to do. He just exhibited his apprehension to the tape.

We received the message from the airborne command post and the MCCC and DMCCC authenticated the message. I recall that the call name for one of the command posts was “Looking Glass”, but am not sure that the order came from this unit. At the defined time, the MCCC initiated the countdown and began the sequence. The countdown progressed normally through the initial phase with no malfunction or indication of anything out of the ordinary. It went as smoothly as one of the many tests we had run in the past using the Launch Signal Responders.

Following the successful completion of the initial phase of the countdown, the DMCCC took his key and moved to level 1 and to the Commit station at that level. The MCCC gave the order to “Commit on my mark-MARK”, and turned his key. The DMCCC obviously turned his key within the allotted time and we began the commit phase. The activities during this phase, as I recall, included final topping of the fuel and LOX tanks, guidance system checks and final setup and transferring of systems to internal control. The initial phase of the commit sequence proceeded normally up to the point at which the launcher platform began its ascent to the surface. One of my responsibilities was to monitor certain areas within the silo during countdown by use of closed circuit TV units. One of the cameras was installed in the missile enclosure area and provided a view from the bottom of the silo toward the underside of the launcher platform after it moved from its lowered position. As the launcher platform rose in the enclosure, I noticed a liquid cascading from the platform and falling into the silo. I notified the MCCC of this and informed him that this was not normal and that it was evidence of a problem. He acknowledged that he understood and that he would keep the missile at the cap by withholding the initiation of the Abort sequence after reaching the up and locked position. Normally, after reaching the up and locked position during a test the Commander would initiate the Abort sequence to cause the missile to lower into the silo, download the oxidizer back to the silo storage tanks and return the system to a stable condition.

It appeared that the liquid which I had observed was fuel, since the only things it could reasonably be were fuel or hydraulic fluid and it appeared to be less viscous than hydraulic fluid. We clearly had a problem which demanded an investigation before we lowered the missile back to a standby position. Because of the high probability that the fluid was of hydrocarbon nature (whether fuel or hydraulic fluid) and that, because of the location of the plumbing which conveys the oxidizer, it was likely that this plumbing had become contaminated. Transfer of the oxidizer through the contaminated system would be an invitation to disaster. We all knew that the mixing of hydrocarbon and liquid oxygen produces a gel material which is very sensitive to shock and will explode with very little provocation. The only way to transfer the oxidizer from the missile to the silo tanks was through this plumbing and we were unwilling to do this unless we were certain that the system was clean.

The EPPT and I were the ones with responsibility to investigate situations of this nature in the silo. The commander directed us to don our emergency gear and to proceed into the silo to see what the conditions were. Neither of us were enthusiastic about this, as the location to which we must go was at the lower levels of the silo, inside the missile enclosure, and we would have to use the stairway to both get down and to return. It would not be safe to use the personnel elevator with the potential for explosive gases in the silo, as the controls for this unit contained open relays which caused arcing during normal operation. These relays were, incidentally, quite troublesome at some site. The contacts tended to burn and weld, disabling the elevator.

At this time, one of the upper level officers (Colonel),an observer at the event, directed that the Non-Essential Motor Control Center be opened. The idea was that doing this would deactivate all of the non explosion proof receptacles in the silo. This would decrease the probability of ignition resulting from arcing within these fixtures. There are a number of other circuits which are controlled by this MCC, including the silo air handling equipment. None of us recognized the full significance of this action and it did seem reasonable to deactivate the open receptacles. I should add here that the situation which we were experiencing was not one covered by any existing emergency condition checklist, and we were making some of the procedures up as we went along.

Wayne and I prepared to enter the silo dressed in our asbestos suits and wearing our Scott Air Packs. Thus equipped, we had some difficulty maneuvering and I recall wondering how we would be able to make our way down and up the spiral staircase. Nevertheless, we proceeded through the first blast door (known as the debris door) and shut it behind us. We moved down the tunnel and had nearly reached the second door (this is the thick one) when the lights began to dim. The generator was slowing - obviously not a normal situation. Wayne and I returned to the LCC, as he was the one who would have to start the other generator and put it on line. We had operated both generators in parallel during the countdown, as both were required for a tactical launch. (I think we could actually get the launcher platform up with just one unit, but we used both for the PLX) Wayne started the other engine and during this time the operational unit began to pick up speed. He paralleled the unit and shut down the faulty unit.

At some point we heard what sounded like a muffled boom from the direction of the silo. I don’t recall the actual sequence, but it could have occurred during our move through the tunnel.

After having restored power neither Wayne nor I was willing to proceed into the silo. There was clearly something serious happening there and we both knew that entering the silo would likely be the last thing either of us did. Fortunately, the commander recognized this and did not direct us to do so.

During this time there were at least two maintenance personnel involved in connecting the missile tank pressurization unit to the missile. This was a unit located on the cap, intended to provide the control required to maintain proper pressures in the fuel and oxidizer tanks during the extended stay on the cap. This was a normal procedure for a situation in which a missile must remain up and locked for an extended period of time. The two maintenance men were up on the L/P making the connections and I was viewing their activities by means of my cap mounted camera when I saw liquid oxygen begin to gush from the main fill and drain line. This line runs up the side of the missile from the vicinity of the engine skirt to a point above the intermediate bulkhead, where it enters the oxidizer tank. As I recall the line is approximately 8 inches in diameter, and could have been larger. The line was flowing full and, if it continued to flow, would dump the entire load of liquid oxygen from the missile tank in a very short period of time. When the maintenance people realized what was happening, they both climbed down from their positions on the platform and fled the scene. One took the truck, an International Crew cab, and the other ran for the main gate. I recall the one running made it to the gate before the one in the truck. At the time it made quite an impression on me, but the distance was not that great and the driver lost quite a lot of time getting to the truck and getting it started.

Shortly after this, I lost sight of anything on the cap. The cloud of oxygen enveloped the camera and most everything else on the cap.

The oxygen spread out around the cap and quite a lot of it must have flowed into the silo, as well. I am sure that a lot of it must have entered the silo in a liquid state, although I have no evidence to support this.

Shortly after this we all heard a fairly large boom, more intense than the first one, and we lost power in the LCC. We could do nothing but wait from that point on. I recall that as I was moving around the LCC I backed into one of the support structures of the LCC and my Air Pack banged against it making quite a loud noise. This caused most of those present to jump and to express great relief when it became clear that the noise had come from a source other than another explosion.

During this time someone in the silo was in contact with a person, via land line, who was stationed on a hill some distance from the site. As I learned later, this person was providing information as to what was visible from his vantage point and which we would not otherwise have known. He stated that there was a fire directly under the missile and that there was smoke issuing from the silo. He also said that at the time of the actual destruction of the missile, the warhead rotated 180 degrees and fell through the body of the missile.

I don’t recall how long it took for the cap area to reach a condition allowing our escape from the LCC, but it was probably not longer than a half hour. The major action subsided relatively quickly, and the topside observer provided us with the information that the condition was safe for exit. It seems that we had some trouble in getting through the security area on our way out, but I am not sure. In any event we went up the stairway and out into the light. We had been cautioned to not touch anything as we left the complex, probably to preserve all evidence as it existed. I ignored this directive and picked up some things as I walked toward the gate. Others may have, as well, but I don’t know.

I cannot recall whether we were taken back to the base in separate vehicles or not, but after arriving we were put in separate rooms and questioned individually. This went on for about an hour, as I recall, possibly longer. There were subsequent sessions during which we were asked to describe the event, and at one such session, there was a legal representative in presence. One area of questioning had to do with training and whether our crew had experienced all of the required instruction and testing. One of the training activities about which they had questions was one which was crew administered - the MCCC presented some information and was to have certified as to our participation. I don’t recall the subject of the training, but failure to have properly performed it could have reflected poorly on the MCCC. I had reason to believe that something about our training session had not been entirely by the book and I refused to answer the questions. This caused some consternation on the part of the investigators, particularly when the attorney advised them that I did not have to respond and that they could not demand that I tell them why I would not respond. After some discussion and following their assurance that they were not out to hang Capt. Ford, I did answer their questions. My comments apparently did not cause him any harm, as he was promoted to Major shortly after the incident.

I am not aware of any formal, official statement regarding the actual cause of the destruction of the complex, but there were unofficial speculations as to what actually happened. I will detail that as well as I recall.

The fundamental cause of the incident was that the drain sequence for the fuel system did not occur as it should have. This sequence is intended to drain all of the fuel in the lines above the disconnect fittings on the launcher platform. These line are full during Commit to provide for continuous topping of the on-board fuel tank up to the time the L/P begins to rise. The drain sequence takes place during Commit, just before the L/P begins its ascent.

The residual fuel in the L/P lines drained out after they disconnected from the fixed portion of the fuel system. This was what I observed falling in the enclosure.

The speculation is that the fuel vapors were drawn into the silo main air handling exhaust plenum located on Level 2, just at the end of the tunnel. When the Non Essential MCC was tripped the fan in this plenum stopped running and allowed the gases to reach an explosive level. Something caused the vapor to ignite and the plenum disintegrated and sent shrapnel flying around Level 2. All of the electrical wiring which connects the silo with the LCC passes close to the plenum as it enters the tunnel. Some of the flying debris cut into the wiring harnesses and produced the signal which caused the main fill and drain valve on the missile to open. Once this happened there was no chance to save the site.

I have heard speculation to the effect that the fires under the missile, on the L/P, could have caused the ignitors in engine supply turbines to “cook off”. Since the fuel and oxidizer supply lines to the turbines would have been closed in preparation for the PLX, the turbines would spin without the resistance provided by the liquid. They disintegrated and sent particles of debris up through the tanks, causing their collapse. There was a report from the topside observer that there was a fire burning on the L/P just below the missile. This would support the theory of “cook off”.

I cannot explain the generator malfunction. It is possible that the quality of the intake air caused an upset in the combustion mixture. Whether that would produce the situation which we experienced or not, I don’t know.

I have considered attempting to secure the official record of the investigation from the Federal Gov’t. through a Freedom Of Information action. I have not done so, but it seems that the files on this incident should have been declassified by now.

Following this incident, the AF issued a revision to the manual describing the PLX procedure which required that the liquid oxygen be replaced with liquid nitrogen in the silo oxidizer storage tanks as a normal procedure in preparation for a PLX. This may have occurred following the next loss. I think one of the losses occurred after this one, but I am not sure. In any event, the losses caused the AF to rethink the PLX procedure and conclude that using liquid oxygen in the procedure presented more hazard than it was worth.

I do not recall that our crew had any extended time away from duty, but I suspect that we did. We resumed duty at some point and continued in a normal rotation throughout the remainder of the life of the system.