In February I was asked by Proceedings to review the Navy’s report of investigation into the USS Iowa (BB-61) disaster of April 1989. The objective was to gain an experienced naval officer’s professional opinion of the investigation report. The objective was not to conduct a second investigation, nor was I to conduct extensive interviews, examine evidence, or in any other way attempt to duplicate the investigation.
The Navy Chief of Information (CHINFO) made available the basic investigation report, the Naval Sea Systems Command (NavSea) technical report, the transcript from a Senate Armed Services Committee hearing, the Navy Investigative Service (NIS) interview package, the Federal Bureau of Investigation (FBI) report, the transcript of the 7 September 1989 press conference that announced the investigation’s findings, and the 5 March 1990 House Armed Services Committee report on the Iowa investigation report. Since it was first released, this material has been available at CHINFO to anyone who asked to see it. Much of the released material has been redacted, which is lawyer-speak meaning that all names and other personal identifying information have been excised from the reports. This makes for awkward reading, but it does not detract from a review effort. Anyone asking to see the report will be given access to a two-foot-high stack of paper, containing the reports enumerated above.
It is important at the outset to remind readers that this is a report of the Judge Advocate General Manual (JAGMan) investigation. A JAGMan investigation is neither a trial nor an indictment. As Rear Admiral Milligan, the investigating officer, stated in a Senate hearing, “The basic report contains my opinions. The opinions I reported are inferences and conclusions drawn from the facts. . . . Regulations require me, to the extent possible, to identify what I perceived to be the cause of the incident. . . . The purpose of a JAGMan investigation is not to convict; it is to discover and learn.”
The Navy has been widely criticized about this tragic incident, most recently by the House Armed Services Committee. Much of the criticism centered on the criminal investigation conducted by the NIS, the spate of leaks during the investigation, and the doubts cast by an inconclusive report from the FBI laboratories after the Navy sent evidence from the explosion to them for analysis. One thing is clear: If you want to know what happened you have to read the entire report, all two feet of it. Skimming the executive summary or reading the transcript of the 7 September press conference does not suffice. In fact, the timing of the Navy’s press conference may have unwittingly contributed to the controversy about this investigation, because the final technical reports, which contain a comprehensive picture, were not endorsed until after the press conference. One frequently heard charge is that the Navy mishandled the investigation entirely—that it went into the criminal investigation mode late in the game, missed critical evidence, and then jumped to conclusions.
A reading of the full report suggests otherwise. Admiral Milligan arrived on board the Iowa one day after the turret explosion. According to his own testimony, he came aboard believing that the explosion had been a tragic accident, and focused at once on shipboard safety, ordnance-handling procedures, manning, training, gun propellant stability, and other possible mechanical issues. He was supported by a technical team from NavSea, under the command of Navy Captain J.D. Miceli, whose main purpose was to find the source of the energy that caused the five bag charges being loaded into the center gun of turret two to ignite. Miceli’s investigation, which was done in support of and parallel with Milligan’s, and whose technical inquiries were also supported by independent Army, FBI, and Navy technical resources, ultimately conducted more than 20,000 tests and analyses.
According to Admiral Milligan, the possibility of a deliberate human act being the cause did not surface until 8 May 1989, three weeks after his investigation had begun. Once it did, he called in the NIS. Thus the overall investigation grew to three parallel efforts: Milligan’s, Miceli’s, and, belatedly, the NIS. In essence, Milligan had to find out what happened, Miceli had to find how it happened, and the NIS had to find out, if possible, who might have done it.
The timing of the reports and the press conference is crucial to understanding some of the controversy. Admiral Milligan signed his initial investigative report on 15 July 1989. The report was then reviewed by the next three levels of command. Commander Naval Surface Forces Atlantic, Commander of the U.S. Atlantic Fleet, and the Chief of Naval Operations (CNO), who signed the report on 31 August 1989. The Navy then held a press conference on 7 September, where it declared that the explosion was caused not by an accident, but by a deliberate human act, and released the initial report of investigation to the press.
Much has been made of the inconclusive FBI report, which seemed to contradict the Navy’s theory about the cause of the explosion. The timing of the investigation contributed to this problem. The report indicates that the Navy originally believed that a timing or detonating device had been put into the powder train. The Navy asked the FBI to analyze materials from the exploded gun, and provided a sample timer-detonator device that was thought to be something an individual could pick up in a local electronics store. The report contains the FBI’s reply, which consisted of two elements. First, it concluded that the timing device theory was not true. Second, it advised the Navy to do more testing, especially materials testing, of the residual evidence of foreign materials found in the gun.
The Navy, in effect, postulated a theory to the FBI, but the FBI could not corroborate it. The report describes how the Navy then did the additional materials analysis suggested by the FBI laboratory, and this time uncovered firm evidence of the presence of a chemical incendiary device. But the sequence of initial report, press conference, and then the final technical report must have contributed to the skepticism with which the Iowa report was met. The possibility of a chemical incendiary device was mentioned in the press conference on 7 September, but it was really nailed down in Captain Miceli’s 27 October report.
Following the press conference, there were charges made that the Navy was covering up the real causes of this disaster— namely, an old ship, old and poorly maintained ammunition, poor training, sloppy supervision, mechanical problems, and more likely ignition sources—in order to protect the battleships from budget cuts.
The technical report, however, contains a convincing account of the tests that proves almost conclusively what did not cause the ignition, and what probably did cause it. The possible accidental causes of propellant ignition in the open breech are listed as follows: unstable propellant, direct flame or spark, a burning ember, a bad primer, frictional heating, impact or compression, electrostatic discharge, hazards of electromagnetic radiation (HERO) radio frequency (RF) spark, or procedural errors. The report then describes the exhaustive efforts of the Navy and outside agencies to ignite the propellant (powder) bags by exploiting any of these potential causes.
According to the report, the Navy tested in detail the possibility that the propellant had been overheated while being stored in a barge. Propellant from the Iowa was compared to baseline samples of the same propellant lot maintained by the naval magazines ashore; the amount of propellant stabilizer remaining in both samples was nearly identical, and above Navy minimum specifications. The investigators also examined the possibility that the propellant had built up an ether-air mixture in the individual powder tanks (cans). The propellant sometimes exudes minute amounts of ether when stored for long periods in the cans; but the problem with this theory is that the bags are detanked below before being sent up to the gun room; if there was any ether-air mixture, tests showed that it was quickly dissipated upon being detanked.
The report revealed that the rammer rod mechanism had traveled 21 inches beyond its normal stop point, so the Navy conducted extensive compression tests on the propellant bags, even running the rammer into a 16-inch gun bore loaded with powder bags at the speed used to seat a 2,700-pound projectile, as opposed to the much slower speed with which a bag of powder is rammed. The worst consequence of that was a ruptured powder bag, which did not ignite. Investigators dropped bags from a 100-foot-high tower; some of the bags broke open, but none ignited.
The Navy also tested extensively the theory of a flame or spark, subjecting the bags to the direct open flame of a cigarette lighter. Bags in their protective covers took almost five minutes to ignite; the raw silk bags went up in three and a half minutes; the black powder ignition bag took nine minutes. The time interval on the day of the explosion, as determined by the investigation, from ramming of the powder to the first explosion, was about 30 seconds.
Then the silk bags were tested for vulnerability to static electricity. It was determined that the maximum static charge that could be built up on people in the turret was zero; on the stationary silk bags it was .5 kilo-volts (kv), and on bags being rammed into the breech, it was .0145–.0185 kv. The testers used a charge of 25 kv on both the igniter pads and the silk bags, but with no effect. They then broke open a black powder igniter pad, exposed the dust, and applied as much as 360 kv, without ignition. It has been suggested that a stray RF field might have set off the powder in the open breech. The bags themselves contain no components that could couple with an ambient RF charge; they are not susceptible to HERO. The primer used in the breech block is susceptible to HERO, but, as the report concludes, three factors argued against primer ignition: no RF source can reach inside the massive steel turret to produce a field; the maximum distance at which a primer can ignite a bag is 28 inches, but the breech face was 42 inches away from the nearest bag, and the steel tray was between the primer and the powder; finally, the primer was found unfired after the explosion.
The investigation did reveal violations of safety rules, use of improper procedures, and deficiencies in documentation of shipboard training. For example, cigarette lighters were found in turret two, despite safety rules prohibiting their presence; the center gun had been loaded with five bags (instead of six) of propellant of a type and in a quantity expressly forbidden in regulations; and propellant had been improperly handled in long-term stowage. Only 13 of 51 watch stations actually manned in turret two requiring formal personnel qualification standards were filled with such persons. Of the 37 positions manned in turret two on 19 April involving power-operated-handling equipment of personnel involved with explosive devices, only nine had current nonnuclear ordnance certification. But the report concluded that none of these directly contributed to the explosion.
The report marches down through the possibilities clearly and carefully, disproving each one. There were no burning embers, since the gun had not yet been fired. Turret one had fired, but the ship was doing 25 knots, and the relative wind would have blown any particles away from turret two.
As the technical report makes clear, never since big guns went to sea in the U.S. Navy have the propellant or the black powder igniters gone off as a result of friction, electrostatic discharge, impact, or crushing. The only explosive incidents on record are the result of a flareback of hot gases from the previous round, or a hot particle left in the chamber from the previous round. In the Iowa case, there was no previous round. Miceli reports that there are many cases on record of torn bags, of propellant grains being caught in the screw threads of the breechblock and crushed, and of loose propellant from a torn bag being thrown into the chamber, but none of them ever resulted in an explosion.
As one reads the report and sees each contending theory of accidental ignition demolished, the obvious question becomes, what did cause it? The picture becomes clear by the end of the report, and the forensic determination of what did cause the explosion makes for fascinating reading, especially the account of how the investigators found the evidence.
When a 16-inch shell is fired, the rear edge of its copper rotating band (cannelure), already engaged by the lands in the barrel after the ramming stroke, is folded over on itself, providing a gas seal so that all the pressure from the exploding propellant acts on the base of the projectile. At the instant of the explosion the copper band, in folding over, creates in effect a time capsule, entrapping the products of combustion from the propellant deflagration into the grain structure of the copper metal.
This is where the laboratories, after being prompted by the FBI report to look deeper, found the telltale molecular evidence in the rotating band of the Iowa's projectile of three materials that had no business being in the chamber of a 16-inch gun. Using such techniques as infrared spectroscopy, X-ray diffractometry, pyrolysis gas chromatography, and mass spectrometry, as well as scanning electron microscopy and energy-dispersive spectroscopy, inductively coupled plasma atomic-emission spectroscopy, and neutron-activation analysis, the investigators discovered molecular evidence of brake fluid, steel wool fibers, and calcium hypochlorite trapped in the metal of the copper rotating band on the Iowa projectile, which remained in the barrel after the open-breech explosion. The labs also discovered the presence of two organic chemicals, phenyl ether glycol and diethyl ether glycol, which are not found in either the propellant or the normal cleaning solutions used in the guns, but which are found in commercial brake fluids and carburetor cleaners. They also found evidence of a heat-sealable mylar polymer and fibers of glass.
The test reports, done by the Naval Weapons Support Center, Crane, Indiana, stated that no evidence of any electronic device was found, and concluded instead that “initiation of the bag charges in the Iowa explosions was caused by a source of intense heat and flames in or between the two bags nearest the projectile. The bags were rammed 21 inches farther in than normal, and were probably still under compression when ignition occurred. Ignition occurred in the black powder igniter pad, not the propellant.” The investigation report describes how the Navy tested such a device in the same circumstances as the Iowa incident using another 16-inch gun. A chemical device, consisting of a plastic bag with steel wool, brake fluid in a glass tube, and calcium hypochlorite, was placed between the first two bags in the chamber of a 16-inch gun, and then compressed by a longer than normal ramming stroke, thus breaking the glass tube. Twenty-five seconds later, the bags ignited. The laboratory then retrieved the same deposits in the rotating band of the test projectile that were found in the Iowa's projectile.
After disproving the many theories of accidental ignition, the technical report concludes that an incendiary device was inserted between the first two bags, the bags compressed by overramming, with an explosion resulting 25–40 seconds later. As Admiral Milligan testified:
“The shipboard program and technical portions of this investigation . . . clearly demonstrate that an accident did not cause this explosion . . . yet an explosion did occur. That is an irrefutable fact. As opinion, I formulated what I then believed and continue to believe was the cause of this explosion . . . a difficult explanation—an explanation that I personally moved toward only with the greatest sense of caution and reluctance. The logic, however, appeared to me as inexorable and relentless: if this tragedy was not an accident, then it must have been the result of an intentional act.
“I had not undertaken this investigation with the expectation that my steps would lead me along such a path or to such a conclusion. I neither sought nor expected to discover a person as an intentional perpetrator. If I expected anything, it was to discover that some accidental or mechanical discrepancy had provoked this devastating calamity. I directed the major portion of my investigative efforts toward the discovery of just such an accidental or mechanical cause. Yet, instead of discovering the expected accident, these efforts eliminated that possibility. I would have preferred otherwise, but my preferences are not dispositive. Now, having reluctantly formulated one difficult opinion, another, equally unpalatable one remained to be recorded.”
The testimony then goes on to detail for the committee those elements of the investigation that address culpability. A reading of the entire report, including the NIS interviews, medical reports, and the technical reports, makes clear what human factors the Navy used to make its conclusions with respect to culpability. A technical support team from the Army’s Aberdeen Proving Ground concurred in the Navy’s technical report that there was no problem with the propellant, that a contaminant was found to have been involved in the explosion, and that the contaminant found could be from an ignition source. The Army’s Ballistic Research Laboratory of the Aberdeen Proving Ground concurred with the technical report about the residue analysis.
The entire report package is somber but highly recommended reading for those who want to restore their confidence that the Iowa disaster was thoroughly and properly investigated.
In the clear light of hindsight, the Navy might have received less criticism if it had done two things differently: first, nothing should have been released or taken to a press conference until all of the technical reports were concluded and reviewed. Second, the report could have stopped in its conclusions on culpability with the opinion that the explosion was caused by a deliberate human act, and let the entire investigation report speak for itself, thereby letting the public and the media form their own conclusions. This opinion reflects some wishful thinking on the reviewer’s part, because, in a JAGMan investigation, the investigating officer is required to offer precisely such opinions, based on all the facts, even if some of them are circumstantial in the sense of legal evidence. The incident described by this report was so horrific that the Navy probably would have been pilloried for stopping short by the same chorus that is now saying that the Navy went too far. At the risk of abusing the point, one cannot get the whole picture without reading the whole Iowa investigation report.
Editor’s Note: Report consists of: Rear Admiral Richard D. Milligan, U.S. Navy, Investigation to inquire into the explosion in Number Two turret on board USS IOWA (BB-61) which occurred in the vicinity of the Puerto Rico operating areas on or about 19 April 1989; report of, 15 July 1989, with endorsements as follows: Commander Naval Surface Forces, Atlantic, 28 July 1989; Commander-in-Chief, U.S. Atlantic Fleet, 11 August 1989; Chief Of Naval Operations, 31 August 1989, and supplemented by Rear Admiral Richard D. Milligan, U.S. Navy Supplemental Report, 9 November 1989. The basic investigation report was supplemented by the Technical Report of Captain Joseph D. Miceli, U.S. Navy, Director Technical Support Team, Naval Sea Systems Command (NavSea-06X), 12 May 1989, with addenda of 19 May, 25 May, 9 June, 22 June 1989, and Final Technical Report, 27 October 1989.