Table of Contents


F. DEFENSE AGAINST STAND-OFF ATTACK.

1. BACKGROUND.

a. Purpose. This section address the facts and circumstances concerning defense against stand-off attack at Khobar Towers. Following the Record Report, additional factual investigation of the adequacy of efforts to protect against the very type of attack that ultimately occurred was necessary. As a result, the investigative team was tasked to answer the following questions:

(1) Did an Air Force "mind-set" cause Air Force commanders at Khobar Towers to fail to appreciate the danger to the perimeter in the face of existing threat and vulnerability assessments?

(2) Did the chain of command direct its focus toward threats to the interior of the base perimeter?

b. Areas to Examine. This section examines what drove the specific "inside - outside" delineation of responsibilities between US forces and Saudi officials at Dhahran. It examines how wing leaders viewed the threat; moving the perimeter fence; actions taken by wing personnel to counter or mitigate the threat; the standards applicable at the time of the bombing; and known limitations associated with the wing’s force protection planning efforts. Analysis of the questions noted above is included.

c. Delineation of Responsibilities.

(1) The specific "inside - outside" delineation of responsibilities between US forces and Saudis at Dhahran is grounded in a number of documents. The Joint Chiefs of Staff Standing Rules of Engagement state that a foreign nation has the principal responsibility for defending US citizens and property within the host country. Joint Pub 3-07.2, Joint Tactics, Techniques, and Procedures for Antiterrorism, recognizes that in preparing antiterrorism plans, "In many situations, through agreement with host nation authorities, the plan will probably evolve into the installation having responsibility ‘inside the wire or installation perimeter’ and the host nation having responsibility ‘outside the wire or installation perimeter.’" Finally, AFPD 31-3, Air Base Defense, says, "Outside the air base TAOR (Tactical Area Of Responsibility), US Army, US Marine, or host-nation forces will have sole responsibility for security requirements."

(2) In addition to this written guidance, principal American and Saudi leaders expressed their understanding of US forces and host nation responsibilities. Lt Gen Carl E. Franklin, USCENTAF Commander, believed the Saudis’ "security was certainly sufficient to protect the compounds that they were guarding" and [referring to exterior security] "the host nation is responsible for security." Major General Sultan Bin Adi Al Mutiri, Saudi Eastern Province commander, viewed the division of responsibility for security within the Kingdom of Saudi Arabia as follows: (1) inside a base, the base commander is responsible for security--a shared responsibility with the coalition forces commanders housed in Khobar Towers, and (2) outside the base, civilian police agencies had sole responsibility. Further, MGen Sultan Bin Adi Al Mutiri said the commander of the Security Wing, Royal Saudi Air Force, who was "dual hatted" as the Director of the Coalition Support Unit, was responsible for coordination of internal security and liaison for external security. General Sultan reinforced this division of responsibility after the bombing when he reminded the American wing commander not to assign any patrol vehicles outside Khobar Towers.

d. Threat.

(1) In the historically volatile Middle Eastern environment, the Kingdom of Saudi Arabia enjoyed a comparatively long period--roughly thirty years--of relative calm. That calm was shattered when a terrorist bomb detonated outside the Office of Program Management Saudi Arabia Air National Guard (OPM-SANG) building in Riyadh on 13 November 1995. This was the impetus for USCENTCOM’s change in the threat level from Medium to High for the Kingdom, as chronicled in AFOSI’s July 1995 and January 1996 Antiterrorism Vulnerability Assessments, respectively. The latter assessment cited a list of five terrorist scenarios for US facilities overseas [Classified material omitted]. Listed in order of potential damage and not in order of likelihood, the scenarios included: [Classified material omitted].

(2) Though members of the wing did not exclude any type or magnitude of threat, testimony supports that their thinking and planning gravitated to a threat on the order of 250 pounds of TNT. The commander and his staff were aware of the size and damage caused during terrorist attacks in Beirut, Oklahoma City and the World Trade Center. But closer to home, the attacks were on a much smaller scale. For example, during the Spring and early Summer of 1996 there were numerous attacks in neighboring Bahrain, but they had been on the order of fifteen pounds of explosives or less. In fact, the necessity of wing commander’s restrictions on personnel travel to Bahrain was questioned by the Ambassador to Bahrain. The notable exception to this small-scale threat was the attack on OPM-SANG. This effort on the part of the terrorists was considered a logistics coup in the respect that neither the Saudis nor intelligence sources believed terrorists capable of smuggling and assembling such a large amount of explosives. The wing commander summed up the threat by indicating that it was general in nature, focused primarily in Riyadh, and that there were no warnings of a device the size of that used at OPM-SANG.

(3) Wing members varied on what they viewed as more probable delivery methods for explosives. For example, the wing commander placed greatest emphasis on preventing the penetration of the Khobar Towers compound. This concern was communicated to the security police commander during the latter’s in-brief with the wing commander on 20 March 1996. In addition to a penetration threat, the support group commander was particularly concerned with a suicide bomber and a satchel charge thrown over the compound fence as possible scenarios to defend against.

(4) With varying emphasis, commanders around the wing expressed other threat concerns, including: standoff weapons including rocket-propelled grenades (RPGs), surface to air missiles, or any other shoulder fired weapons; sniper fire; kidnappings, assassinations and highjackings; sabotage carried out by third country nationals; aerial attack; a car bomb detonating outside the complex’s northern barrier; and a penetrating car bombing similar to what the US Marines had experienced in Beirut, Lebanon in 1983. Focusing on the high rise buildings along Khobar’s northern perimeter, the security police commander described his challenge as preventing the kind of major structural damage that collapsed the Marines’ building in Beirut and the federal building in Oklahoma City.

e. Extension of the Perimeter: Wing leadership recognized the vulnerability from beyond the existing perimeter and took the following steps to address it:

(1) The support group commander asked his Saudi counterpart, Colonel Qahtani, to expand Khobar’s perimeter by moving the concrete barriers a short distance outside the fence. The request also included moving the fence and clearing the vegetation growing on the north perimeter to improve visibility. The request regarding the east and west perimeters was granted and concrete barriers were placed approximately five feet outside the fence. The requests to remove the vegetation and move the fence and place the concrete barriers outside the north perimeter were denied. According to the support group commander, Colonel Qahtani explained that expanding the perimeter would encroach on an existing parking lot which serviced a city park and a mosque. The security police commander and the AFOSI Detachment Commander were also rebuffed by their Saudi counterparts when they asked to expand the perimeter a short distance. The likelihood of moving the fence was expressed in most doubtful terms by several Americans.

(2) When it appeared the Saudis would not allow the perimeter to be expanded, wing officials took another approach. They sought to have the parking lot closed–the Saudi response was no. The next request was for the Saudi police to patrol the area just outside Khobar Towers. After repeated requests, Saudi officials continued to assure their counterparts that the patrols would begin and increase to a level sufficient to satisfy the concerns. Over time, wing officials verified that Saudi police patrol coverage increased to twenty-four hours a day.

(3) The Saudis also agreed to establish a method for wing security officials to conduct license plate checks of vehicles viewed as suspicious around the American compound. However, the Saudis did not respond to requests for information in a timely manner. There were often no follow-up checks, leaving US security officials "skeptical" of the program’s value.

(4) Al-Khobar civilian police initiated undercover police operations outside Khobar Towers’ north perimeter. The Saudis repeated their assurances of external protection, guaranteeing they could lock down the entire city (Dhahran) in five minutes. Responding to concerns expressed by US personnel about possible surveillance in the parking lot north of the compound, the Officer in Charge of the Red Beret, Royal Saudi Land Force Military Police said, "We’ve go (sic) patrols out there. We’ve got undercover police. We’ve got sources watching that area. It’s not a problem at all." The combination of: Saudi actions outside the fence; their repeated assurances that they had security outside Khobar Towers taken care of; the perceived threat; and the security enhancements the wing had produced specifically designed to counter the perceived threat, led the wing leadership to delay further efforts to move the fence.

(5) The wing commander did not broach the subject of moving the fence with his counterpart for a number of reasons. First, the threat was not perceived to require it. Second, he was encouraged with what he viewed as improving Saudi police performance. And third, given the force protection improvements being accomplished throughout the wing, the commander did not believe extending the fence was required.

f. Wing Security Enhancements: Wing security enhancements included:

(1) Increasing security police presence by adding additional posts when the wing transitioned to THREATCON BRAVO immediately following the Riyadh bombing.

(2) Placing concrete "jersey" barriers around the Khobar Towers complex as depicted in photo 1 below. A second row of barriers was later added around the perimeter of the complex which served to harden the perimeter against a penetration attempt and to help mitigate the blast effects of a bomb detonated at the perimeter. Discussions between the security police commander and the wing’s explosive ordinance disposal flight led leaders to believe that while an OPM-SANG size device could blow out all the windows of the buildings along the perimeter, such a blast "would not bring down one of those buildings." The support group commander also believed the jersey barrier placement provided "reasonable protection given the threat that we had." Many of these barriers were subsequently spiked down to the street after a vehicle driven by an unknown individual pushed one of the barriers with his car.

Photo 1 - Khobar Towers Northern Perimeter

(3) Reinforcing those portions of the fence line adjacent to streets with trash dumpsters to offer additional protection against high speed penetration.

(4) Repairing damaged portions of the fence line around the complex.

(5) Installing concertina wire along the top of the perimeter fence to preclude personnel access as illustrate in photo 2 below.

(6) Moving jersey barriers away from the perimeter fence to preclude an intruder from using it as an assist to step over the fence as shown in photo 2 above.

(7) Installing two rows of concertina wire on the ground along the outside length of the compound’s perimeter.

(8) Increasing security police manning at static posts. This included building two sandbagged defensive fighting positions to use as M-60 machine gun positions at the Khobar Towers main gate.

(9) Increasing security police vehicle and foot roving patrols inside the compound.

(10) Posting security police on top of towers within the Khobar Towers complex to provide an over-watch or early warning function. Equipped with binoculars, night vision devices, and a land mobile radio, security police observers represented the wing’s best defense, in terms of early warning, against a sniper or other standoff threat. Once these posts were established, the security police squadron commander conducted scenario-based training with posted security policemen to ensure they understood their mission–particularly the rules of engagement associated with their duties. These training sessions consisted of an exchange of ideas based on "what if" scenarios posed by the commander. He extended these training sessions to include security force members assigned to M-60 machine gun positions, as well as the one and a half ton truck ramming posts. In addition to threat detection, they were also able to monitor and assess the patrolling patterns of Saudi civil and military police outside the compound. Notations of Saudi police patrol frequency were made by security police rooftop observers. These data were used by the security police commander to seek additional Saudi support if patrol frequency declined. Observations recorded and maintained on top of building 131 were not recovered after the bombing.

(11) Creating two mobile ramming posts where security police used two and a half ton trucks as a blocking force for gate runners.

(12) Implementing 100% identification checks at compound entry point.

(13) Initiating 100% inspection of all contract and non-coalition vehicles at compound entry point.

(14) Initiating random inspections of coalition vehicles.

(15) Increasing the total number of military working dog (MWD) teams (shown in photo 3 below) at Khobar Towers from two to six.

(16) Initiating random MWD inspections of all parked vehicles inside the compound and of all non-coalition vehicles entering the compound entry point.

(17) Installing jersey barriers to form a serpentine approach along access roads to limit speed on ingress and egress as shown in photo 4 below.

(18) Pruning back vegetation along the northwest perimeter to improve visibility outside the compound.

(19) Denying all vehicle access to garages within the compound.

(20) Securing jersey barriers in place along the perimeter by staking them to the road with pin posts.

(21) Blocking service roads which provided vehicle access to the spaces between individual buildings within the compound.

(22) Altering travel patterns for wing personnel to avoid unnecessary exposure or to prevent lucrative targets.

(23) Increasing security police patrols between Khobar Towers and King Abdul Aziz Air Base.

(24) Initiating discussions within the Wing Security Council to control the movement of third country nationals at King Abdul Aziz Air Base. This issue was still being worked at the time of the 25 June bombing.

(25) Instructing and reminding all wing personnel to be suspicious and inquisitive about strangers, to inspect the interior and exterior of buildings for suspicious packages, and to increase examination of all mail for letter or parcel bombs. The battle staff directives (BSDs) directing the increased vigilance yielded almost instant results. On 16 November 1996, for example, a wing member reported finding a suspicious package at the military post office. Another suspicious package was found, and dealt with, the evening of 29 January 1996 on the ground floor of the Army dormitory.

(26) Wing leaders received feedback from visitors concerning the visible improvements to security. The Consul General from Dhahran, for example, remarked that it appeared the wing commander had "created Fortress America," prompting him (Consul General) to wonder what may be overlooked at his own compound. Likewise, Major General Kurt B. Anderson, JTF SWA commander, said that a good deal of improvement [security enhancements] was evident every time he visited Khobar Towers. He specifically mentioned increases he observed in observation posts, barriers at the front gate, as well as barriers used to block internal portions of Khobar Towers.

g. The Attack. Shortly before 2200 local, 25 June 1996, an explosive-laden truck penetrated the defensive measures established by the Saudis and parked against the northern perimeter fence of the Khobar Towers complex. Moments later the truck exploded, killing 19 Americans, injuring hundreds more American and Saudi citizens. The size of the bomb was later estimated as the equivalent of 20,000 pounds of TNT. After the bombing, the majority of forces were moved out of Khobar Towers.

2. STANDARDS, REGULATIONS AND INSTRUCTIONS.

a. DoD D 0-2000.12, DoD Combatting Terrorism Program, 27 August 1990. THREATCON BRAVO measure 14 required, "Where possible, cars and objects such as crates, trash containers, etc, are to be moved at least 25 meters from buildings, particularly those buildings of sensitive or prestigious nature."

b. DoD 0-2000.12-H, Protection of DoD Personnnel and Activities Against Acts of Terorism and Political Turbulence, February 1993, offered general security considerations for new construction at DoD Sites. Chapter 9 included suggested stand-off distances for buildings. Paragraph B.4.b.(1), states, "100-foot minimum setback between perimeter and building exterior whenever possible." For existing facilities, DoD 0-2000.12H states that an unobstructed area or clear zone should be maintained on both sides of perimeter physical barriers. The inside clear zone should be 30 feet and the outside clear zone should be 20 feet.

c. AFPD 31-3, Air Base Defense, March 2, 1995, established policy for base defense operations, and in the context this report, addressed: Air Force components’ of joint commands responsibilities in ensuring force protection; and outlined installation commanders’ responsibilities for the defense of assets under their control. Following are the specific references:

(1) Para 2.3 "USAF components of joint commands will provide planning support to ensure adequate forces and intelligence are dedicated to protect USAF resources. The Air Force component, in coordination with the joint command, will determine defense force latest arrival times and tailor defense force structure and equipment to match force protection needs for Air Force warfighting resources."

(2) Para 2.4 "Installation commanders are responsible for the defense of assets under their control. The installation chief of security police is normally the defense force commander (DFC) and will plan and execute base defense operations. The DFC will lead those forces provided by the installation commander and other defense forces in the air base TAOR."

d. AFI 31-209, The Resource Protection Program, November 10, 1994, set requirements for the physical security of Air Force personnel, installations, operations, and assets, and identified the requirements of the Resource Protection Program (RPP).

(1) Of the four primary objectives of the RPP highlighted in the this AFI, two applied to perimeter defense. These included: maintaining the Air Force war fighting capability by reducing damage to Air Force resources and safeguarding Air Force property by reducing the opportunity for theft or terrorist attack by making a potential target inaccessible or unattractive.

(2) Specific wing level responsibilities contained in this AFI are limited to Chapters 1 and 2, Responsibilities and Program Management, respectively, and apply only to the installation commander. Following are the applicable references:

(a) Para 1.7. "Develop and implement either an installation RPP (IRPP) or installation security plan (merging is acceptable)."

(b) Para 2.5.6. "Installation Entry Point Checks. The installation commander determines when, where, and how to implement random checks of vehicles or pedestrians."

e. AFI 31-210, The Air Force Antiterrorism (AT) Program, July 1, 1995, established responsibilities and guidance for the Air Force Antiterrorism Program, and in the context of this report, provided guidance on how to establish a local Antiterrorism Program. Specific wing level responsibilities contained in this AFI are contained in Chapter 2, Responsibilities, and apply to the installation commander.

(1) Para 2.14. "Installation Commanders. Establish an antiterrorism program, tailored to local mission, conditions, and terrorist threat using the publications listed in attachment 1." Note: Attachment 1 referenced in para 2.14 above lists twenty-three separate and various DoD, Joint, Air Force, and Army directives, regulations, instructions, publications, handbooks, and guides.

(2) Para 2.15.2. Commanders At All Echelons, "Plan, train, exercise, and execute antiterrorism measures as specified in DoDD 0-2000.12, where appropriate. Each organization implements physical security procedures to protect against terrorism by installing physical security equipment, implementing THREATCONs, employing RAMs, and responding to terrorist acts."

f. The wing published the wing Installation Security Plan to execute its force protection mission. Provisions of the plan included aspects of physical security, law enforcement, resource protection, antiterrorism, and base defense. The plan established the wing’s antiterrorism program addressing the mission, conditions, and terrorist threat. In the context of preparing for perimeter attack, the plan’s following annexes applied: tasked organizations; intelligence (including a terrorist threat appendix); operations (including entry control procedures and measures addressing THREATCON ALPHA, BRAVO, CHARLIE, AND DELTA).

g. The office of primary responsibility for USCENTCOM Regulation 190-2, USCENTCOM Antiterrorism Measures, July 12, 1995, indicated that this publication was "written primarily for security assistance organizations" and was intended for USCENTCOM units only. This publication was not implemented in the 4404th Wing (P).

3. KNOWN LIMITATIONS.

a. In the context of preparing for a perimeter attack, the wing’s force protection planning efforts were hampered by several limitations. These included:

(1) No Written Agreements. The effects on US forces operating overseas in countries without written agreements with the host nation, particularly no status of forces agreement, is detailed in the Section II-C.

(2) Inability to Operate Off Base. In the context of force protection efforts, US forces were prohibited from freely operating outside of their compound–Saudis had the sole responsibility. This division of responsibility--Americans and Saudis inside the fence, Saudis outside the fence--produced an unavoidable seam in the overall defensive posture.

(3) Higher Headquarters Assistance. While the wing hosted many visits from general officers in its chain of command and their functional staffs, higher headquarters support to the 4404th Wing (P) in the form of traditional inspections or staff assistance visits were limited to a visit from Headquarters AFOSI, a USCENTAF Civil Engineer Staff Assistance Visit (SAV), and a USCENTAF Fire Protection SAV. The wing had not been subjected to any formal inspections.

(4) Available Intelligence. While intelligence information was available to security police forces to use in their defense plans, it lacked specificity in two critical areas: location and magnitude of a likely attack.

4. ANALYSIS.

a. Did an Air Force "mind-set" cause Air Force commanders at Khobar Towers to fail to appreciate the danger to the perimeter in the face of existing threat and vulnerability assessments?

(1) No. Existing DoD publications clearly spell out the delineation of responsibilities between US and host nation forces vis-¢ a-vis "inside" or "outside" the perimeter responsibilities. This division of responsibility applies to all services, not just the Air Force. This division of responsibility was well understood at Khobar Towers; it was clear that the Saudis had responsibility for force protection outside the perimeter and shared responsibility inside the perimeter.

(2) At Khobar Towers, Air Force commanders appreciated the danger from a stand-off attack in the face of existing threat and vulnerability assessments. Wing leadership, as well as most people in the theater, generally viewed their threat in terms of an explosive device comparable in size to the device which terrorists had detonated at OPM-SANG in November 1995. They also developed a list of likely scenarios terrorists could employ to deliver such a device. These two elements–the size of the expected bomb and the possible delivery methods–formed the basis for the antiterrorism measures carried out by the wing.

(3) Antiterrorism measures were implemented in accordance with the 4404th Wing (P) Installation Security Plan. Provisions of the plan included aspects of physical security, law enforcement, resource protection, antiterrorism, and base defense. The plan established the wing’s antiterrorism program and was tailored to its mission, conditions, and its assessment of the terrorist threat, satisfying the antiterrorism measures specified in DoDD 2000.12. It should be noted that the measures listed in DoDD 2000.12 were suggested measures based on the nature of the threat. In the context of preparing for a stand-off attack, the plan’s annexes specified, among other things, responsibilities for intelligence (including a terrorist threat appendix) and operations (including entry control procedures and measures addressing THREATCON ALPHA, BRAVO, CHARLIE, AND DELTA). This plan also satisfied the requirements set forth for the installation commander in AFIs 31-209 and 31-210. Namely, its existence satisfied the requirement for the installation commander to implement an installation security plan (AFI 31-209, para 1.7.).

(4) The combination of increasing patrols, adding an additional four military working dog teams, adding a rooftop watch, and increasing security police roving vehicle and foot patrols provided the wing added vigilance both inside and outside the perimeter fence. Although assigned "inside" the wire, these troops were mobile and capable of making, noting, and reporting observations "outside" the wire. Likewise, increasing the manning at static posts, pruning back the vegetation along the perimeter, and increasing the travel of security police patrols between Khobar Towers and King Abdul Aziz Air Base, directly contributed to increased security awareness inside and outside the compound.

(5) In addition to preparations for a stand-off attack, efforts were made to extend the perimeter. Wing leaders made several requests to the Saudis to extend the perimeter. While the Saudis agreed to move concrete barriers outside the east and west perimeter of Khobar Towers approximately five feet, they refused to grant the wing’s requests to expand the north perimeter a short distance. Americans then sought permission for the Saudis to close the parking lot located north of the Khobar Towers complex. This request was also denied by Saudi officials. Instead, the Saudis agreed to: increase civil police patrol coverage to twenty-four hours a day; check license plates of suspicious vehicles as requested by American security officials; and initiate undercover police operations outside Khobar Towers’ north perimeter. Every expression of concern about the north perimeter vulnerability made by wing officials received repeated assurances by the Saudis that they were responsible for, and in control of, the area outside Khobar Towers.

(6) The stand-off distance on the northern perimeter of Khobar Towers was approximately 80 feet. This included a hard distance of approximately 60 feet between the buildings and the perimeter fence, and another 20 feet, covered with fairly heavy vegetation, from the fence to the curb in the parking lot. On 25 June 1996, the tank truck parked at this curb; the bomb crater was centered approximately 80 feet from the front of building 131. This standoff distance met the suggested guidance of DoD 0-2000.12-H for existing buildings and was consistent with the stand-off distances for other US occupied buildings in the AOR.

b. Did the chain of command direct its focus toward threats to the interior of the base perimeter?

(1) Yes. This was their primary responsibility based on the division of responsibility between the 4404th Wing (P) and the Saudi government. This focus was not, however, at the expense of appreciating the danger to the perimeter. Efforts undertaken by the wing to harden Khobar Towers from a penetration attack were extensive. Encircling the entire perimeter of the compound with a row of concrete jersey barriers (later doubled), served as the centerpiece for their anti-penetration efforts. Many of these barriers were subsequently spiked down to the street after wing members discovered a barrier had been displaced by a vehicle driven by an unknown individual. At the same time, these barriers served to deflect blast from a stand-off weapon.

(2) The jersey barriers were complemented with a variety of other physical security enhancements. For example, the concertina wire strung along the length of, and on top of, the perimeter fence served to increase the difficulty for personnel intrusion. Repairing damaged portions of the fence and moving jersey barriers away from the perimeter fence also served to harden the perimeter against personnel infiltration. Denying all vehicle access to garages within the compound and blocking service roads which provided vehicle access to spaces between individual buildings within the compound eliminated the ability of a would-be bomber to deliver an explosive device next to, inside, or under a building within the compound.

(3) Jersey barrier placement also helped distance the interior of the compound from a satchel charge threat. Additional jersey barriers, as well as dumpsters, were also strategically placed to limit speed on ingress and egress and to reinforce areas thought to be vulnerable to high speed approaches. Building two M-60 machine gun positions and creating two mobile ramming posts where security police used two and a half ton trucks as a blocking force for gate runners, capped the wing’s anti-penetration efforts.

(4) Security measures taken at the compound’s entry point were also extensive. These included: conducting random, and frequent, inspections of coalition vehicles and instituting 100% identification checks. The wing commander implemented these checks through his battle staff directives, satisfying the requirement in AFI 31-209. In addition, all contract and non-coalition vehicles were subjected to a complete vehicle inspection, as well as a military working dog explosive check prior to obtaining entry to the compound. Made possible through USCENTAF support, these efforts were designed to discourage or disrupt the efforts of any would-be saboteur.

(5) Again, the command’s focus was not exclusively turned inward. For example, through numerous battle staff directives, the leadership took steps to educate the wing’s population on the hazards of off-base travel. Designed to lower their individual profiles, the wing commander’s travel security directives were aimed at increasing personnel security by lowering the probability of victimization through kidnapping or assassination.

c. Preparations for Perimeter Attack.

(1) Preparations for a perimeter attack undertaken by the wing, while extensive, were geared toward a threat of an explosive device of roughly 250 pounds in size, the size anticipated by all involved. The device, estimated in the 20,000 pound range, far exceeded what the wing had prepared for based on the threat assessment, and far exceeded the size of any device ever before witnessed in the region. As for specific preparations, when wing leaders were unsuccessful in their efforts to have the perimeter fence moved, they pursued other options with the Saudis. Beginning with failed requests to close the parking lot, wing members persisted until assurances were made by the Saudis to provide 24-hour police patrol coverage, check license plate numbers of suspicious vehicles requested by American, and provide undercover police coverage in the area.

(2) Wing officials also intensified their efforts to protect buildings along the compound’s northern perimeter. These efforts included doubling the jersey barriers which was calculated to mitigate the blast effects of a device the size of what they believed to be the threat. Again, in the context of blast effects, the security police commander was concerned about preventing the kind of major structural damage capable of collapsing their apartment buildings in the same fashion witnessed during the bombings in Beirut and Oklahoma City. The wing’s explosive ordinance disposal input during blast mitigation discussions led leaders to believe that while an OPM SANG size device could blow out all the windows of the buildings along the perimeter, such a blast would not bring down one of those buildings.

(3) Based on what he thought the nature and size of the threat to be, the wing commander did not consider redistributing wing members who lived in exterior rooms (closest to the perimeter) to interior rooms (farthest from the perimeter), which would be more vulnerable to a penetration bomb exploding inside the compound. Likewise, the wing commander considered mylar treatment for windows, but deferred it to the wing’s five-year plan. As discussed in detail below, he believed other force protection measures the wing had implemented allowed this postponement.

(4) The wing obtained tactical high-ground by creating security police observation posts on the rooftops of buildings strategically located along Khobar’s perimeter. These posts were designed to address a number of concerns. Most obvious was their early warning capability in detecting threats external to the compound. Rooftop observers were able to visually assess all avenues of approach to the compound for both vehicular and pedestrian traffic. Equipped with binoculars, night vision devices, and a land mobile radio, security police observers represented the wing’s best defense, in terms of early warning, against a sniper or other standoff threat. It was from one of these rooftop posts (top of building 131) that a security policeman detected the truck bomb at approximately 2150 local, 25 June 1996. And as mentioned earlier in this report, these observers also provided route surveillance--an "overwatch" capability--for US forces traveling between Khobar Towers and King Abdul Aziz Air Base. In addition to threat detection, they were also able to monitor and assess the patrolling patterns of Saudi civil and military police outside the compound. Notations of Saudi police patrol frequency were made by security police rooftop observers. The security police commander used this information to seek additional Saudi support if patrol frequency declined. Observations recorded and maintained on top of building 131 were not recovered after the bombing.

(5) Given the perceived threat, the scale, magnitude, and progress of physical security enhancements undertaken throughout the wing, and the continuous reassurances of protection outside the compound by the Saudis, the wing commander’s efforts to protect the perimeter were reasonable to protect against the type of attack that ultimately occurred.

5. CONCLUSIONS.

a. The 4404th Wing (P) was in compliance with the applicable Air Force instructions detailing requirements for resource protection and antiterrorism security. Force protection efforts against a stand-off attack were reasonable based on the known and expected threat.

b. Wing leaders protected those areas over which they had responsibility to control. The magnitude and scope of on-going force protection efforts illustrates that Air Force commanders at Khobar Towers did not fail to appreciate the danger of a stand-off attack in the face of existing threat and vulnerability assessments.

c. [Classified material omitted].

d. Air Force commanders at Khobar Towers did not fail to appreciate the danger to the perimeter in the face of existing threat and vulnerability assessments.

e. The chain of command did focus toward threats to the interior of the base perimeter, but not at the expense of appreciating the danger to the perimeter.

G. EVACUATION PLANNING, PRACTICE AND EVALUATION.

1. BACKGROUND.

a. Purpose. This section addresses the facts and circumstances concerning evacuation planning, practice, and evaluation at Khobar Towers. Specifically, the investigative team has been tasked to answer the following questions:

(1) What were the evacuation plans in place at Khobar Towers on 25 June 1996?

(2) Did those plans take into account efforts to minimize injury from an explosive blast?

(3) Did prior building evacuations occur in five minutes and, if so, why were only three of ten floors of Building 131 evacuated on June 25 in approximately the same time?

(4) Were the prior evacuations sufficient as substitutes for planned evacuation drills?

b. Khobar Towers Compound.

(1) The 4404th Wing (P) has occupied the United States portion of Khobar Towers since June 1992. Khobar Towers was built by the Saudi Government in 1979 and remained essentially unoccupied until it was used to house the almost 3,000 US military personnel of the wing, along with British and French military personnel.

(2) Khobar Towers is approximately one kilometer east of King Abdul Aziz Air Base. The US military housing area occupies a small portion of the Saudi housing compound of approximately 14 city blocks. The US military section is bordered to the northeast by civilian housing, to the southeast by Saudi military housing, to the southwest by a large vacant lot, and to the northwest by a public park and parking lot. The complex is located in the midst of an urban environment, surrounded by residential and commercial areas and mosques.

(3) The US military compound--approximately 40 buildings--is in an area roughly the size of two city blocks. Four of these buildings are used by the British and French Coalition Forces. The buildings range in size from four to eight stories.

(4) The buildings that suffered the most damage as a result of the 25 June 1996 bombing are located at the north end of the compound, buildings 131 and 133. These dormitory-type structures are eight story "T" shaped structures constructed of 5 ½ inch precast, concrete wall panels and 6 ½ inch floor panels. The exterior wall contained approximately 18% glass. The facilities, when originally built, did not have any fire protection devices such as evacuation alarms, sprinkler systems or smoke detectors. Smoke detectors were later installed in all occupied areas of the dorms. A complete discussion of building alarm systems is included below in section II H, Communications.

c. Building Evacuation Planning. There were two separate emergency response procedures for Khobar Towers in place on 25 June 1996. The first procedure was an evacuation procedure used to respond to bomb threats and fire. The second procedure, outlined in the emergency response plan and contained in the dormitory attack response procedure, was used to respond to warning of an enemy attack. The night of 25 June 1996, the security police posted on the roof of Building 131 reacted to the truck as a potential bomb threat and initiated evacuation procedures. If they had viewed this as an enemy attack, then they would have initiated the attack response procedures.

(1) Bomb Threat/Suspicious Package and Fire Evacuation Procedures.

(a) In the event a suspicious package or smoke or other fire indicators were discovered, the individual discovering the event would report it to central security control (CSC) or the fire department. The CSC or fire department would then direct the individual reporting the event to initiate an evacuation of the building. A security police patrol or fire response team was then sent to the scene. Additionally, CSC could request Giant Voice activation by first contacting the chief of security police who would call the wing operations center (WOC) who would then get permission from the wing commander to activate Giant Voice. In the event of a suspicious package, the arriving patrols would set up a 300 foot cordon. To evacuate additional buildings within the 300 foot cordon, a security police patrol would be dispatched to the building, use the elevator to go to the top floor and begin knocking on doors, notifying the occupants to evacuate. As individuals left their rooms, they were instructed to notify other occupants to evacuate as they worked their way down and out of the building. This method of evacuation notification was described as a waterfall method.

(b) Individuals would evacuate the building through the stairwell and proceed to their designated squadron assembly area for personnel accountability. These procedures were developed in April 1996 after the AFOSI detachment commander suggested to the wing commander that evacuation plans for the Khobar Towers facilities be re-accomplished to include evacuation procedures in the event of a terrorist attack. On 13 April 1996, at a meeting concerning the terrorist threat, each squadron was tasked to develop individual evacuation plans for its facilities. These were submitted to the civil engineer readiness flight to consolidate and deconflict the assembly points. A review of all the plans collected after the bombing indicates that evacuation plans were submitted in April 1996, coordinated and deconflicted, and returned to the squadrons. Plans for 21 of 34 buildings within the Khobar Towers compound were available for review. Plans for buildings 131, 133, 129, and 127 were included in the submissions. Additionally, a review of the plans indicated at least one unit distinguished fire evacuation procedures from terrorist evacuation procedures by changing the assembly point–perimeter for fire and interior of the compound for terrorist attack. However, most units did not make a distinction. A copy of these new plans was provided to each of the squadrons and posted on the inside of the doors of each room.

(2) Enemy Attack Response Procedures.

(a) The wing developed standardized emergency procedures for an enemy attack. These procedures were used to meet the threat of a SCUD attack during the Gulf War. In the event of an incoming attack, Giant Voice would be activated by the wing operations center (WOC) in the siren mode, followed by voice instructions that an attack was probable (Alarm Yellow) or imminent (Alarm Red). Giant Voice is a base-wide loud speaker system with speakers located on selected roofs of the Khobar Tower buildings and throughout the American portion of King Abdul Aziz Air Base. A further discussion of Giant Voice is contained in Section II-H, Communication.

(b) After a Giant Voice warning that an attack was imminent, the Dormitory Attack Response Procedures directed individuals to seek shelter in the front hallway, the inner most area of their suite. The procedure specifically instructed them to remain away from the kitchen, television room, and individual bedrooms because these rooms were located to the outside of the building. The procedures were published in 1994 and posted in the dormitories. The attached example sheet was recovered from one of the rooms in Building 133 after the bombing on 25 June 1996.

(3) Training. Personnel assigned to live in Khobar Towers received a briefing on general evacuation procedures during the Right Start Program from senior wing officials. Briefings on the specific evacuation plans for each dormitory were the responsibility of each squadron. Squadrons posted the instructions on bulletin boards and on the backs of their room doors.

d. Evacuation Practice.

(1) There was no formal program within the wing to practice evacuation exercises. However, from November 1995 through June 1996, there were 14 suspicious package incidents within Khobar Towers. These incidents were reported in the security police blotters. In three incidents, the blotters had specific entries noting building evacuations. Circumstantial evidence indicates some of the other incidents may have also resulted in evacuations. For example, on 18 November 1995 a suspicious vehicle was parked adjacent to Building 111. The Security Police blotter indicates a 300 foot cordon was established, but it does not specify that any buildings were evacuated. Applying a 300 foot cordon around building 111 indicates that buildings 109, 110, 112, and 113 should have been evacuated. The security police blotter indicates this incident was initiated at 1509 and at 1518 evacuation of all personnel from the affected area was complete. Thus, it appears that the evacuation took nine minutes from the time the suspicious package was first reported to CSC until evacuation of personnel was complete.

(2) A review of the 14 suspicious package incidents, indicates that a 300 foot cordon was implemented in six of the incidents. In each of these incidents, the establishment of the cordon should have driven building evacuations. This is consistent with witness recollections of about six evacuations during this period. A copy of the Khobar Towers Plot Plan was used to determine which facilities fell within the 300 foot cordon around the effected facility. Using the premise that the Security Police conducted an evacuation of all facilities within the 300 foot cordon, building 131 would have been evacuated twice during this period. The chart below indicates building evacuation and timing information available for each incident:

Suspicious Package Incidents Nov 95 - Jun 96

Date of Incident

Time

Initiated

300’ Cordon

Buildings Evacuated

Time Evacuation

300’ Cordon

Complete

Time Incident Terminated

Time Required to Complete Evacuation

14 Nov 95

1350

X

108, 109, 110, 111, Garage 12

unknown

1453

unknown

16 Nov 95

0936

X

not

Khobar Towers

unknown

1101

unknown

18 Nov 95

1509

X

109*, 112*, 113*

1518

1522

9 min

14 Dec 95

1218

X

127*, 129*, 130*, 131*, 132*

unknown

1225

unknown

29 Jan 96

1935

X

115*, 117*, 118*, 119*, 120*

unknown

1950

unknown

3 Feb 96

1817

no cordon

n/a

n/a

1822

n/a

13 Feb 96

1245

no cordon

n/a

n/a

none listed

n/a

14 Feb96

2155

no cordon

n/a

n/a

2240

unknown

16 Feb 96

0652

no cordon

n/a

n/a

none listed

n/a

8 Mar 96

1304

no cordon

101

1306

1315

2 min

27 Mar 96

1050

no cordon

n/a

n/a

none listed

n/a

6 May 96

1950

no cordon

n/a

n/a

2010

n/a

9 May 96

1700

X

129, 127*, 130*, 131*, 132*

1706

1836

6 min

14 Jun 96

2105

no cordon

n/a

n/a

2114

n/a

*No entry in blotter that building evacuated, but within the 300’ cordon

(3) Fire Protection Evacuations. The fire protection unit did not schedule or conduct any fire evacuation exercises, but they encouraged facility monitors to conduct drills. Emergency evacuation procedures were discussed during the mandatory Right Start inprocessing briefing for all new personnel. Additionally, the fire department performed internal exercises for proficiency training of its personnel which were conducted on a monthly basis during April, May and June 1996. Each time a different facility within Khobar Towers was used. One of these drills was in the civil engineer facility, building 133. These fire training exercises tested the fire department’s capabilities of fighting a fire in a high-rise facility or office complex. During these exercises fire department personnel evacuated the facility before conducting the exercise. However, they did not time or evaluate the evacuation of personnel. Of note, the British coalition members within the Khobar complex conducted monthly fire evacuation exercises.

e. Evacuation Evaluations. As noted above, the wing had no program to evaluate the conduct of evacuations. The only noted evaluations of any programs related to evacuations were two USCENTAF Staff Assist Visits (SAVs), a fire protection SAV in January 1995 and a civil engineer SAV in January 1996. Neither SAV identified any deficiencies related to evacuations by the fire department. Further, although the SAVs identified a number of deficiencies in the civil engineer readiness flight, none were related to evacuation planning.

2. STANDARDS, REGULATIONS, AND INSTRUCTIONS.

a. Contingency Planning and Evacuation.

(1) Joint Publication 3-07.2, Joint Tactics, Techniques, and Procedures for Antiterrorism.

(a) Joint Pub 3-07.2, by its own terms, applied to commanders of combatant command, subunified commands, joint task forces, and their component commands. Thus, it applied to USCENTCOM, USCENTAF, and the 4404th Wing (P). Joint Pub 3-07.2 procedures were to be followed except in exceptional circumstances.

(b) Joint Pub 3-07.2 reaffirms the basic responsibilities of command, but also provides specific procedures. Drills and exercises are mandated as the best test of an installation’s ability to respond to a terrorist incident short of an actual event. Guards provide early warning and detection and special orders should address, at a minimum, their response to an approach by unauthorized personnel.

(c) In the event of a bomb threat in a building, the Joint Pub 3-07.2’s directed response was search, movement within the building, partial evacuation, and total evacuation. Evacuation procedures were to be prepared, publicized, and rehearsed in advance. Also to be addressed in the plan are alarm systems, assembly areas, routes to to assembly areas, personnel evacuation response, building and area clearance, and evacuation drills. The alarm system should be easily distinguished from the fire alarm. Routes to assembly areas should not approach the bomb at any time. Evacuation drills were to be periodically practiced under the supervision of the installation or unit senior office.

(d) As to enemy attack, Joint Pub 3-07.2 contains no specific procedures beyond calling for drills and exercises.

(e) Above the 4404th Wing (P) Commander level, USCENTAF and USCENTCOM had command responsibility. As Joint Pub 3-07.2 states, "every commander, regardless of echelon of command or branch of Service, has an inherent responsibility for planning resourcing, training, exercising and executing antiterrorism measures to provide for the security of the command. Theater Combatant Commanders are to "Ensure proper coordination of all local policies and measures for protecting DoD facilities, resources, equipment, personnel, and family members in foreign areas from terrorist acts and for assisting their subordinate commanders in implementing Military Service programs." These combatant commanders were also to designate a staff office, usually the J-3 or law enforcement or security section, to supervise, inspect, test, and report on the base antiterrorism programs within theater.

(2) AFI 31-210, The Air Force Antiterrorism (AT) Program, 1 Jul 95.

(a) AFI 31-210 establishes the Air Force’s antiterrorism program. Major commands are required to establish an antiterrorism program tailored to the local mission, conditions and the terrorist threat. This was in addition to their oversight responsibility exercised through their command security police chief under AFPD 31-2, Law Enforcement, 6 May 94.

(b) Installation commanders are required to establish an antiterrorism program tailored to local mission, conditions and terrorist threat. They are directed to use DoD 0-2000.14-H, Protection of DoD Personnel and Activities Against Acts of Terrorism and Political Turbulence, and Joint Publication 3-07.2, Joint Tactics, Techniques, and Procedures for Antiterrorism. Installation commanders are also required to ensure the installation can respond to a terrorist attack. AFI 31-210 required commanders at all echelons to plan, train, exercise and execute antiterrorism measures as specified in DoD 0-2000.12 where appropriate. Commanders are also responsible for integrating lessons learned from actual incidents and operation exercises, to correct deficiencies. The installation chief of security police has the functional responsibility for implementing these requirements.

(3) AFI 32-4001, Disaster Preparedness Planning and Operations, 6 May 94.

(a) Requires each Air Force Installation to have a disaster preparedness plan to include an annex on enemy attack. There is no requirement for the plan to address evacuation. Similarly, training of installation personnel in disaster preparedness responses is required.

(b) Under AFI 32-4001, the 4404th Wing (P) Commander was required to conduct attack response exercises twice a year as the installation was located in a chemical-biological threat area. These exercises were to provide "realistic, integrated, large-scale training for the installation and the DRF [Disaster Response Force]." It also directs the conduct of a major accident response exercise (MARE) at least once every quarter. It requires the Installation Commander to establish an Exercise Evaluation Team (EET). The EET develops exercise scenarios and debriefs and critiques performance and prepare a report to track discrepancies.

b. Fire Protection. Air Force Policy Directive (AFPD) 32-20, Fire Protection, requires fire drills be conducted by referencing and adopting the requirement set by the National Fire Protection Association (NFPA) Code 101, Code for Safety to Life In Buildings and Structures. Fire exit drills must be conducted with sufficient frequency to familiarize all occupants with the drill procedures and to have of the drill a matter of established routine. There is no established time standard in which fire drills or actual fire evacuations must be completed.

3. LIMITATIONS.

a. 90 Day Rotation. The 90 day rotation increased the number of personnel to be trained and the frequency necessary to ensure all were trained in evacuation procedures. Approximately 10% of the wing’s personnel turned over every week. There would be approximately 200 to 300 new arrivals each week. At this rate of turnover, for example, only 30% of the personnel living in Khobar Towers on the 25 June 96 would have been present for the 9 May 1996 evacuation of several buildings.

b. Saudi Restrictions on Exercises. Conduct of wing exercises was limited to King Abdul Aziz Air Base and inside the Khobar Towers compound. Exercises were limited so as not to offend the Saudis. These restrictions were never written down but expressed verbally to commanders. Specific Saudi restrictions included using the Giant Voice system in the siren mode and not being able to use the primary crash net for exercises. The Saudi restrictions are evidenced by the last minute cancellation of the mass casualties exercise in December 1995 and the Saudi reluctance to allow the wing to test the Giant Voice siren after the 25 June 1996 bombing.

c. Chemical Gear Requirements. The readiness flight within the civil engineer was also responsible for the chemical defense program. They spent a significant amount of time working chemical gear requirements. They were required to complete a 100% check of chemical gear for each individual arriving in the AOR, 200 to 300 people each week. When 100% checks of chemical gear were started in the fall of 1995, only 20% of personnel arriving in the AOR had all required gear. As a result of the readiness flight efforts, by June 1996, this figure had improved to approximately 90%. The effort expended in ensuring chemical gear requirements were met distracted the readiness flight from completing other responsibilities.

d. 4404th Wing (P) Mission. The wing’s flying mission in support of the "no fly" sanctions was a 7-day a week, 24-hour a day operation. Exercises in the Khobar Towers compound were limited to avoid disturbing the crew rest (a 12-hour non-duty period with 8 hours of uninterrupted rest before a flying mission) of aircrews. Additionally, most personnel assigned to the wing worked significantly longer during their 90 day temporary duty than when assigned to CONUS units. Finally, the provisional wing was not staffed for all the functions of a permanent Air Force wing.

e. Readiness Personnel. The civil engineer readiness flight had four authorizations. In the two weeks before the bombing, the flight was only 50% staffed as a result of early rotation of personnel. Additionally, although the most recent readiness flight chiefs were trained, several previous readiness flight chiefs had no training or experience in readiness. This was identified in the USCENTAF Civil Engineer Staff Assistance Visit in January 1996 and had been corrected for the rotations after April 1996.

4. ANALYSIS.

a. Evacuation Planning.

(1) Responsibility for Evacuation Planning. Within the wing, each squadron was responsible for establishing procedures for evacuation of their facility. The readiness flight within the civil engineer squadron assumed the responsibility to review the plans and were in the process of trying coordinate and develop a wing plan. They also had the responsibility to develop the attack response emergency plans. These responsibilities are outlined in AFI 32-4001. The Readiness Flight coordinated on the April 1996 rendition of evacuation plans submitted by the squadrons.

(2) Initiation of Emergency Response Procedures. On the night of 25 June 1996, security police were posted on the roof of Building 131. Their posting required them to specifically watch and note any unusual occurrences on the northern and eastern perimeters. If they perceived any unusual occurrences they were required to notify Central Security Control (CSC). They observed the Saudi fuel truck back up to the fence on the north perimeter and immediately notified CSC. Based on the nature of the perceived threat they characterized it as a bomb threat; and initiated immediate evacuation procedures.

b. Evacuation Procedures In Place at Khobar Towers. On 25 June 1996, two emergency response plans were in place at Khobar Towers.

(1) Evacuation. Upon notification to evacuate, all personnel were to proceed down the stairwell to their squadron area for accountability. Notification was via the "waterfall" method of starting at the top floor and going room-to-room to notify personnel to evacuate. On 25 June 1996, notification was made by the three roof-top security policemen who observed the truck pull up to the northside fence. These procedures had been updated in April 1996 and posted in Khobar Tower buildings. These procedures met the requirements in Joint Publication 3-07.2, as implemented by Air Force Instruction 31-210. They were written, described escape routes, and included procedures to account for all personnel after evacuation. These evacuation procedures had been previously used to respond to seven suspicious packages which necessitated evacuations from November 1995 to June 1996. The evacuation procedures did not specifically take into account efforts to minimize injuries from an explosive blast. These evacuation procedures were not designed to minimize injuries from blast, they are used to quickly remove individuals from the vicinity of a perceived threat, including a suspected bomb.

(2) Although the evacuation procedures employed did not specifically take into account minimizing the effects of explosive blast, the difficulty of designing a plan to account for multiple possibilities is demonstrated by applying the following possible timing sequences to the bombing at Khobar Towers.

(a) Bomb exploding approximately three minutes after notification to evacuate. On 25 June 1996, the top three floors had evacuated and most personnel who were notified were in the stairwell when the bomb went off. This provided them some protection from the effects of explosive blast.

(b) Bomb exploding 12 minutes after notification to evacuate. If the bomb had exploded 8 minutes later, 12 minutes after notification to evacuate, it is likely personnel would have been outside but close to the building with little or no protection from explosive blast.

(c) Bomb exploding 20-30 minutes after notification to evacuate. If a timing device similar to the one used in the OPM-SANG bombing had been used, there would have been 20 - 30 minutes from the time the Security Police saw the truck and the bomb explosion. This would have allowed for a 300 foot cordon to have been set up and complete evacuation of buildings 131, 130, and 129 away from the area. This would have minimized injuries from explosive blast.

(3) Dormitory Attack Response Plan. The second procedure in place was the Dormitory Attack Response Plan. This was a written procedure for enemy attack, developed in December 1994 and posted on the doors of each room in the Khobar Tower facility. The attached copy of the procedure was recovered from building 133 after the bombing. These procedures were implemented through the wing Survival Recovery Center plan. Personnel were directed to take cover until the attack was over. Accountability was reported to the dormitory manager. These procedures also met Air Force requirements for an emergency response plan. They were written, described emergency actions, and included procedures to account for all personnel after the attack. By directing personnel to seek shelter in the inner most area of the suite, away from the exterior of the building, the Dormitory Attack Response Plan took into account efforts to minimize injuries from an explosive blast.

c. Previous Evacuations. As discussed in the background section, the security police blotters indicate that there were 14 suspicious package incidents between November 1995 and June 1996. The blotters indicate that 3 of these suspicious packages were treated as bomb threats and resulted in building evacuations. Four additional incidents have a blotter entry indicating a 300 foot cordon was setup. This is consistent with security police quick reaction checklist (QRC) #14 which required a cordon be setup for a bomb threat. QRC #14 also requires evacuation of all the buildings within the 300 foot cordon. Thus it appears from an analysis of the blotters that there were seven bomb threat incidents which resulted in building evacuations. This number of evacuations is consistent with the support group commander’s recollection of 6 to 10 evacuations as a result of suspicious packages.

d. Timing of Previous Evacuations. The support group commander estimated that the suspicious package evacuations were all completed within five minutes. He based this estimate on information he received from his fire chiefs. Evidence available to assess this estimate includes security police blotter entries, the on-scene fire chief from the 9 May 1996 evacuation, and several other witnesses who were present during the evacuation on 9 May 1996.

(1) Security police blotters. Of the seven incidents which resulted in building evacuations, the security police blotters indicate timing for three events. On 18 November 1995, the blotter indicates all personnel were evacuated from within the cordon nine minutes after the incident was reported. This entry does not specify how many buildings were evacuated. Three buildings should have been within the 300 foot cordon. Thus it appears it took nine minutes to evacuate the three buildings on this occasion. The blotters also indicated an evacuation of building 101 on 8 March 1996. The security police blotter indicates this took two minutes. There is no indication that a 300 foot cordon was set up for this incident.

(2) On 9 May 1996, the security police blotters indicate that building 129 was evacuated in six minutes. Evidence indicates that a total of five buildings were evacuated on this occasion, including building 127. At least one witness who was evacuated from building 127 estimated it took 10 to 15 minutes to evacuate. Additionally, another witness estimated that it took about ten minutes to evacuate the eight story buildings. Finally, the fire chief who was present and on scene for the 9 May 1996 evacuation opined that it was not possible to evacuate one of the larger facilities within 5 minutes.

(3) There is no evidence to indicate that any procedures were established to accurately assess the time it took to evacuate facilities during these incidents. As noted above, the only written records that could be used to determine any evacuation times comes from the security police blotters; they were not consistently annotated with evacuation times. Additionally, testimonial evidence indicates that it took longer five minutes to evacuate the buildings.

(4) Summary. The evidence does not support that prior evacuations were completed within five minutes. Evidence indicates prior evacuations of buildings in Khobar Towers complex were completed in 10-15 minutes. This is consistent with time it took to evacuate the first three floors the night of the bombing.

e. Prior Evacuations - Adequacy as Substitute for Exercises.

(1) The investigative team was also tasked to review whether the building evacuations for suspicious packages were sufficient substitutes for planned evacuation drills. They were not.

(2) The basic Air Force contingency exercise program is contained in AFI 32-4001, Disaster Preparedness Planning and Operations. This instruction establishes the Major Accident Response Exercise (MARE) program. A MARE is any one of various accident scenarios and must be conducted at least quarterly. MARE scenarios frequently require the establishment of a cordon and the evacuation of personnel and buildings within the cordon. AFI 32-4001 also requires an Attack Response Exercise be conducted twice a year in chemical threat areas, including Saudi Arabia. Further, the Antiterrorism program requires exercises to evaluate the effectiveness of the base antiterrorism program.

(3) Under the Air Force exercise program, each exercise is evaluated by the wing exercise evaluation team (EET). The chief is appointed by the installation commander. The EET establishes exercise objectives and provides a debriefing, critique, and report for each exercise. Additionally, if the EET identifies a discrepancy during the exercise, the EET must also track concomitant corrective actions and complete a trend analysis. The purpose of exercises is to determine the installation’s and assigned units’ capability to respond to operate during, and recover from combatant and noncombatant contingency operations. Additionally, in AFI 31-210, commanders at all echelons are required to integrate lessons learned using after-action reports from actual incidents and operational exercises, to correct deficiencies.

(4) Using these guidelines then, a real-world event may be an adequate substitute for an exercise, as long as the real-world event is used to identify potential deficiencies and integrate lessons learned. Wing officials indicated they were confident in their procedures because of the previous real-world evacuations. However, no documentary evidence was found indicating the wing used the real-world evacuations to evaluate their response program and identify any areas requiring improvement. Additionally, because there was no accurate method used to time the evacuations, wing officials did not have a reliable estimate of lead-time necessary to evacuate a facility. This is particularly true of the large tower facilities. Finally, with the high-rate of turn over of personnel, the frequency of these real-world events may not be enough to properly ensure personnel were aware of procedures to follow in the event of an evacuation. For example, there were no evacuations between 8 March and 9 May 1996. During this time, almost 60% of personnel in the wing would have turned-over.

f. Fire Drills. Fire drills were not conducted. The standard for conducting fire drills is in the National Fire Protection Association , Code 101, Code for Safety to Life from Fire In Buildings and Structures. This code does not specify an interval to perform the drills; but it does require them to be conducted frequently enough for everyone to be familiar with the procedure and to have the conduct of the drill "be a matter of established routine." The suspicious package evacuations performed within the wing did not meet this requirement.

5. CONCLUSIONS.

a. Evacuation and attack response procedures were in place at Khobar Towers on 25 June 1996.

(1) Evacuation plans were updated in April 1996. They were not designed to minimize injuries from an actual explosive blast. They were designed to remove individuals from the vicinity of a perceived threat as quickly as possible. If time had permitted, people would have been evacuated.

(2) Dormitory attack response procedures were developed in December 1994. These plans require individuals to take immediate cover in the interior of the building in the event of an imminent attack. As such, they minimize the risk of personal injury from explosive blast.

(3) The security police personnel on the roof of building 131 made a decision to initiate an evacuation rather than an attack response. Such action was reasonable based on the uncertainty of the perceived threat.

b. Prior evacuations of buildings in Khobar Towers occurred in approximately 10 to 15 minutes. This is consistent with the time it took to evacuate the top three floor of building 131 on 25 June 1996. There are no set time standards for the evacuation of buildings.

c. The wing was required to practice emergency evacuation procedures, but did not do so. Although the wing had a number of real-life evacuations from November 1995 to May 1996, these real-life evacuations were an inadequate substitute for exercises. The lack of specific data collection during the real-life evacuations left the wing with an inflated estimate of its capability to evacuate the buildings in Khobar Towers. The lack of an accurate estimate of building evacuation times did not impact the wing’s response on 25 June 1996.

d. The wing did not have procedures in place for the practice of fire drills as required by the National Fire Protection Association Code 101.

H. COMMUNICATIONS: (1) AUTOMATED BUILDING ALARM SYSTEM, (2) RADIO LINKS BETWEEN SENTRIES AND THE GIANT VOICE SYSTEM AND (3) FULL-TIME INTERPRETERS.

PURPOSE. This section provides an expanded factual analysis of communication capabilities and deficiencies as they existed at Khobar Towers on 25 June 1996. Specifically, it addresses:

a. the lack of an automated building alarm system

b. the lack of direct radio links between sentries and the Giant Voice system

c. the absence of full-time interpreters, and whether that impeded rapid communications with Saudi security forces and police

AUTOMATED BUILDING ALARM SYSTEM.

1. BACKGROUND.

a. US military personnel stationed within the vicinity of Dhahran were housed in a number of high rise buildings located within a fourteen city block residential area in a suburb of Dhahran called Al Khobar. The US compound was comprised of approximately 40 high rise and other buildings within the northwest two city blocks of the Al Khobar complex. In addition to housing personnel, the area also contained the normal range of service and support functions found on other US military facilities. US Army soldiers were also billeted in the US compound, as were British and French coalition forces, who occupied four buildings in the southwest corner of the compound.

b. The Downing Assessment stated " . . . Saudi construction standards for Khobar Towers-type buildings did not require a fire alarm system. Consequently, US forces moved into facilities that did not have a system that could have served for mass warning notification of an attack." Use of a building fire alarm system as a general automated building alarm system for attack warning violates the provisions of Joint Publication 3-07.2, which notes the distinction between the two types of alarms and cautions that the "bomb threat system should be easily distinguished from the fire alarm."

c. The wing relied on a Giant Voice system, with siren and voice capability, for attack notification. This system is common to most operational Air Force wings. The system at Khobar Towers had been installed during the Gulf War, primarily to warn of Scud missile attacks. Because bomb threats are localized, the wing used a "waterfall" notification system to alert people to evacuate building(s) and areas affected by the particular threat. The Giant Voice system and the wing’s evacuation and attack response plans are discussed more specifically in Section II G.

d. Saudi building codes did not require the installation of a fire alarm system. Smoke detectors were installed in the sleeping areas of the US occupied buildings. Smoke detectors were installed instead of an automated facility alarm (fire alarm) system largely because of the small amount of flammable items contained in the buildings and the fact that the buildings were constructed predominantly of prefabricated cement sections and other masonry products. The civil engineering squadron commander, who was also the base fire marshall, and his deputy concurred with earlier opinions that the buildings were not combustible. In fact, the fireball resulting from the bomb did not result in a fire.

e. The British coalition forces living at Khobar Towers occupied buildings equipped with fire alarm systems.

f. The lack of an automated fire alarm system was brought to the attention of senior staff members on several occasions by the wing fire chiefs. The fire chiefs were virtually unanimous in their opinion that the lack of a centralized fire alarm system was not a problem. Three stated that, based on their experience, an automated fire alarm system was not needed at Khobar Towers and that the wing did not need to spend the money on one. Their reasoning was also due to the very few flammable items in the facilities, the most flammable being the air conditioners. This belief was further supported by the fact USCENTAF, during two civil engineering staff assistance visits, did not identify the lack of a fire alarm system as a deficiency. With this information in mind, the support group commander and wing commander did not see the requirement for a near-term project. However, they did not want to lose sight of the improvement and, hence, provided the planning wedge in the outyears of the five year plan.

g. With the evolution of an overall wing facilities plan, the purchase and installation of an automated fire alarm system was included in the wing’s Five-Year Facilities Improvement Plan, called Vision 2000, dated 24 May 1996. The project for a centralized alarm system was programmed for FY00 with an estimated cost of $250,000.

h. The basic premise of a fire alarm system is to provide a quick warning and have people expeditiously exit the building to a safe place. This differs from an attack warning which (1) warns people to seek refuge in the interior of facilities away from glass windows, (2) to close the doors as they leave their rooms, and (3) to await the notification of the "all clear," or other instructions.

i. In the event of the need to immediately evacuate a building for fire or a bomb threat, the individual discovering the event would report it to central security control (CSC) or the fire department. The CSC or fire department would then direct the individual reporting the event to initiate an evacuation of the building. Evacuation was completed by taking an elevator to the top floor and knocking on doors, notifying the occupants to evacuate. As individuals left their rooms, they were instructed to notify other occupants to evacuate as they worked their way down and out of the building. This method evacuation notification was described as the waterfall method.

2. STANDARDS, REGULATIONS AND INSTRUCTIONS.

a. AFI 32-2001 and DoD Instruction 6055.6 required compliance with Military Handbook 1008B, Fire Protection for Facilities Engineering, Design, and Construction, 15 Jan 94. AFI 32-2001, The Fire Protection Operations and Fire Prevention Program, 15 May 1994, applies to personnel who develop and implement fire protection and fire prevention at Air Force installations, facilities, and contractor-operated facilities. Para 3.5, Fire Engineering, states that the fire chief provides operational expertise and also requires the engineering flight chief to manage fire engineering as prescribed by Military Handbook 1008, Fire Protection for Facilities Engineering, Design, and Construction. Additionally, provisions of DoD Instruction 6055.6, DoD Fire and Emergency Services Program, 15 Dec 94, required facility compliance with Military Handbook 1008B. Para E.2. requires DoD Components’ fire and emergency services programs to include the elements described in Enclosure 2 to DoDI 6055.6.

b. Military Handbook 1008B establishes protection engineering policy and criteria for DoD components and applies to DoD facilities located on or outside of DoD installations, whether acquired or leased, or third party financed and constructed. The Handbook incorporates by reference National Fire Codes fire protection criteria published by the National Fire Protection Association (NFPA). National Fire Codes include NFPA Standard 101, Life Safety Code Para 2.1.1 of the Handbook requires that building construction related to building egress and safety must comply with NFPA Standard 101. NFPA Standard 101 contains requirements for fire alarm systems. For existing dormitory and apartment buildings, para 1.3.2 of the Handbook does not require modification of the buildings to meet more stringent Handbook 1008B requirements, so long as the buildings are acceptable to the "authority having jurisdiction" and they meet NFPA 101 standards. The sole "authority having jurisdiction" for the Air Force is the Chief Fire Engineer of the Air Force, HQ AFCESA/DFE, at Tyndall AFB.

c. Chapter 17 of NFPA 101 contains general fire protection requirements for existing buildings occupied as dormitories. Paragraph 17-3.4.1 requires installation of a fire alarm system. Paragraphs 17-3.4.3.1, 7-6.3.2, and 7-6.3.4 require the evacuation notification to occupants to be an internal audible alarm that meets the NFPA’s National Fire Alarm Code.

d. Joint Pub 3-07.2, Joint Tactics, Techniques, and Procedures for Antiterrorism, (25 June 1993), Appendix K, paragraph 1f, requires bomb threat procedures to address an alarm system. Paragraph 1g notes the distinction between bomb threat systems and fire alarm systems, the "bomb threat system should be easily distinguished from the fire alarm."

3. ANALYSIS.

a. The decision not to install a fire alarm system was based on the construction of the facilities and the opinions of numerous fire chiefs and headquarters fire protection specialists that such a system was not required. Based on these inputs, the wing commander, support group commander and civil engineering squadron commander deferred the planned upgrade, programming the $250K cost for a fire alarm system into the wing’s first Five-Year Facilities Improvement Plan.

b. If more personnel had simply evacuated to the outside of their buildings, there may have been additional casualties as evidenced by the fact local Saudis located approximately 800 feet from the blast site received serious injuries.

c. In the instant case, had a fire alarm system been activated, security police and fire department emergency response units could have been reacting to building 131 when the explosion occurred. Members of these units could have been exposed to the blast, as well as building occupants responding to a fire, versus evacuation alarm. A fire alarm system should not be used as a substitute for a bomb threat warning system.

4. CONCLUSIONS.

a. Fire alarm systems were not installed based the professional opinions, experience, and recommendations of various fire protection experts. In light of the buildings’ construction, low incidence of flammable material and the availability of smoke detectors, the wing commander, as did preceding commanders, did not believe the installation of a fire alarm system was immediately required.

b. Joint Publication 3-07.2 explicitly cautions against any procedure that may confuse fire and other types of alarms. Implication that a fire alarm system might be used as a general automated building alarm system are not in line with this directive.

c. There is no DoD requirement for an automated building alarm system designed and installed for other than fire warning purposes, i.e. attack or bomb threats.

d. For bomb threat warning and notification, the alarm system planned and relied on by the wing employed the "waterfall" process. While not automated, it was an effective notification method and satisfied the requirement of Joint Publication 3-07.2 for an alarm system in the event of a bomb threat.

e. For attack warning, the wing relied on its installed, tested and operational Giant Voice system, which is specifically designed to provide base-wide notification of impending or imminent attack. This system is common to virtually all Air Force installations at both CONUS and overseas locations, and well known to Air Force members.

f. On 25 June 1996, the security police perceived the tank truck as a potential bomb threat, vice attack, and initiated the appropriate notification and evacuation procedures. While the activation of an automated alarm system (not fire) for the building would most likely have had more people moving toward exits, the short time between recognition of the threat and the explosion makes any meaningful estimates of survivability unlikely.

DIRECT RADIO LINK BETWEEN SENTRIES AND THE GIANT VOICE SYSTEM.

1. BACKGROUND.

a. On 25 June 1996 the land mobile radio communications system at Khobar Towers and King Abdul Aziz Air Base (KAAAB) included the wing command net and the security police communication networks. The wing command net fell under the operational control of the wing operations center (WOC) and central security control (CSC) controlled the security police net. This arrangement was a legacy from Desert Storm and is similar to that found at most Air Force installations.

b. The WOC was located at KAAAB approximately four miles from Khobar Towers while the CSC was located at Khobar Towers. During normal day-to-day activities, the WOC provided the wing’s emergency actions response and reporting, was the primary interface for the wing’s senior commanders, and was the primary point of contact between the wing and higher headquarters. During emergencies and contingencies, the WOC was staffed by senior organizational representatives to advise the battle staff (wing commander and his senior staff) of operations within their respective units. The CSC was connected to the WOC by telephone and radio.

c. The Giant Voice siren was the means used to notify base personnel of an impending enemy or Scud missile attack and was controlled by the WOC. Procedures for the security police to activate the Giant Voice system were as follows:

(1) observation post/patrol requests CSC to activate the alert system

(2) CSC requests the WOC activate Giant Voice

(3) WOC notifies the wing commander of the request seeking his permission to activate Giant Voice

(4) if permission received, the WOC activates Giant Voice to warn KAAAB and Khobar Towers of an emergency.

d. Security police used a quick reaction checklist for the activation of Giant Voice. The checklist is a variation of a well known protocol for activating Giant Voice. Though procedures may vary slightly from base to base, these procedures preclude the inadvertent activation of the Giant Voice siren which may unnecessarily alarm the civilian and base populace.

e. Security police sentries and patrols did not have direct access to the WOC or to the command net. They communicated with the CSC and between one another via the security net. WOC personnel monitored the security police net on a continuous basis.

f. On 25 June 1996, the security police commander and his personnel were not allowed to activate Giant Voice. This was in accordance with wing procedures and their quick reaction checklist that required the security police sentries to notify the CSC who would then notify the WOC, which had sole authority for activation of Giant Voice. This checklist is a variation of a well known protocol for activating Giant Voice. Though procedures may vary slightly from base to base, they are designed to preclude the inadvertent activation of the Giant Voice siren which may unnecessarily alarm the civilian and base populace. Shortly after the bombing the wing commander adjusted procedures to authorize the chief of security police, his operations officer or the senior security police manager to directly activate Giant Voice.

2. STANDARDS, REGULATIONS AND INSTRUCTIONS.

a. There are no regulations or instructions for the connectivity of radio nets or the activation of Giant Voice.

b. AFI 32-4001, Disaster Preparedness Planning and Operation, 6 May 1994, provides only general guidance concerning warning and notification systems.

(1) "7.1. Warning and Notification. Every Air Force installation must have a rapid and effective system to disseminate disaster information quickly. These alarm signals are for passive defense applications: do not use them for active ground defense warning and notification."

(2) "7.1.1. Use signals that are compatible with local, host-nation, or theater systems. Follow command or theater guidance when more than one warning and notification system could apply."

3. LIMITATIONS. Giant Voice is effective in the siren mode but is difficult to understand in the audio mode. As an outdoor audio system, it was difficult to hear in enclosed rooms, especially over the sounds of air conditioners, televisions or music. The system tended to draw listeners to the windows to hear better, which may place personnel in harms way.

4. ANALYSIS.

a. At issue is whether the multi-tiered procedure for activating the Giant Voice system was appropriate. Since there are no objective standards providing direction or guidance on the employment of a Giant Voice system, it is necessary to analyze the reasonableness of the commander’s decision to provide this multi-layered process for activation. Under all the circumstances, it appears that the system in place was reasonable.

b. The most significant limitation is in the system itself. It is difficult to hear. Messages must be spoken very slowly to be understood. As noted in the limitations, it draws personnel to the window in order to hear the message. As such it is only appropriate for use in certain circumstances not requiring the most urgent response, such as assuming mission oriented protective postures well in advance of attack or in announcing all clear. The wing operations center could use radios and telephones, as needed, to supplement Giant Voice.

c. Another important factor is the political environment. The Saudi government was sensitive to public displeasure at having a large foreign military force stationed in one of its major cities. It was thought that frequent use of the Giant Voice would alarm the local citizens and exacerbate the situation. Thus, the US forces, in deference to the Eastern Province Commander, would have minimized the use of Giant Voice to those circumstances consistent with the host nation’s desires.

d. Since Giant Voice is not particularly suitable to urgent cases, and the political concerns mitigated against its use except for limited circumstances, a system carefully controlling its use was appropriate.

5. CONCLUSION. The decision concerning the procedures for Giant Voice activation, voice or siren, is left to the discretion and considered judgment of the commander. Limited authorization to activate the system, used to warn of attacks, was reasonable under the circumstances.

FULL-TIME INTERPRETERS.

1. BACKGROUND.

a. From the time the wing moved to Dhahran in 1991, it relied on coordinating with the Saudis through those host nation contacts who spoke English. Shortly after his arrival, Brig Gen Schwalier observed the difficulties associated with not having a full-time interpreter and hired the wing’s first, and only one, in September 1996. Wing senior leadership were of the belief that the interpreter did an outstanding job interfacing with the Saudis.

b. The interpreter was assigned to the support group commander, but the majority of his time was spent supporting the security police. In addition to assisting the security police, the interpreter worked on various claims and assisted the AFOSI detachment. In the event of competing requests for his services, the support group commander would set priorities. When the interpreter was on vacation, the wing returned to its old methods of coordination, relying on English speaking Saudi liaisons until his return.

c. The interpreter was provided a pager and cell phone for easy contact. When security police saw what they believed to be suspicious activity outside the fence line, they paged the interpreter and requested he coordinate their concerns with the proper host nation forces. The Saudis required US authorities to contact their Saudi military police counterparts, who would in turn notify the civilian police. The system was not timely, and many times the suspicious person had departed the area before host nation forces arrived. The night of the bombing, he was off-base at a local mall and was paged by security police. He immediately contacted them via his cell phone. Security police advised him that a fuel tank truck had just backed up and parked near the bushes along the northern perimeter, and they needed him to call the Saudi military police as soon as possible. The interpreter attempted to immediately call the Saudi military police (the Red Hats), but their telephones were busy. He then heard the bomb explosion.

(1) Security Police Support. Prior to the wing hiring the interpreter, the security police had limited communications with the Red Hats who had security responsibilities at Khobar Towers. After the interpreter was hired, the security police used the wing interpreter at will–night and day.

(2) AFOSI Support. AFOSI mostly relied on English speaking Saudis, but in those cases where an interpreter was required they used the wing interpreter. They also used the services of the Protocol Officer at the US Consulate, Dhahran and a US Army Military Intelligence linguist for introductions and translations with host nation officials.

2. STANDARDS, REGULATIONS AND INSTRUCTIONS. There are no standards for assigning or hiring interpreters for wings.

3. ANALYSIS.

a. Brig Gen Schwalier was the first wing commander assigned for one year, and the first to recognize the limitations imposed by the lack of an interpreter and to do something about it. He hired the first interpreter for the wing and assigned him to the support group since its members had the most frequent contact with the Saudis, especially the security police.

b. The assigned interpreter was on-call 24-hours a day and did an admirable job for one individual. Other than those times he was on leave, he always responded when paged, called or requested. However, one interpreter was hard pressed to satisfy all the needs of the wing. In fact, even the wing commander recognized that the wing would have benefited from an additional interpreter, but efforts to hire another interpreter were not initiated before 25 June 1996.

4. CONCLUSIONS.

a. There are no standards for assigning or authorizing interpreters for wings. As in this case, wing commanders must rely on their own judgment and resources to fill the position. That is what Brigadier General Schwalier did.

b. On the night of 25 June 1996, the interpreter had been contacted and was seeking to establish contact with the Saudi military police when the explosion occurred. Given the wing’s reaction time from assessment of a potential threat to explosion, approximately three to four minutes, and the fact the interpreter was attempting to call the Saudi military police, indicates that the notification was proceeding in a timely manner. Saudi procedures for notifying their military police who would notify their civil police were being followed. The availability of the one full-time interpreter was not a factor in the wing’s response to the suspicious truck.

c. There is no indication having only one interpreter affected coordination with Saudi officials on 25 June 1996.