6
Comparison of Pulsed Fast Neutron Transmission Spectroscopy and FAA-Certified Explosives-Detection Systems

In this chapter, PFNTS is compared with the existing FAA-certified EDSs on the basis of their practicality for improving aviation security. All existing FAA-certified EDSs are based on x-ray CT, and all deployed certified EDSs are either InVision CTX-5000 or CTX-5000 SP systems. The data on PFNTS detection capability were provided to the panel by Tensor Technology (testing at the University of Kentucky) and the University of Oregon, and all of the PFNTS operational characteristics evaluated by the panel are based on Tensor Technology's MDNR design. The panel's intention, however, is not to compare Tensor Technology and InVision Technology but to make as comprehensive a comparison as possible between PFNTS and x-ray CT systems in terms of their acceptability for deployment in airports.

Performance

The motivation for deploying explosives-detection equipment is to improve aviation security. The selection of the equipment is determined by its ability to detect explosives and, at the same time, cause minimal disruptions of airline operations. The FAA's certification requirements reflect both of these requirements (FAA, 1992). Although the specifics of the FAA's certification requirements are classified, the factors that are measured are not. Perhaps the most important factor is the Pd (probability of detection). To pass certification testing, an EDS must be able to detect various explosives configurations at a rate determined by the FAA. Because a high Pfa (false alarm rate) could impede airline operations, the FAA has also set a requirement for a minimal Pfa. Finally, the FAA requires a throughput rate of 450 bags per hour.

The detection of explosives is the fundamental performance criterion for comparing competing technologies. The results of certification testing of the InVision CTX-5000 are classified, and some results from blind tests of the Tensor MDNR cannot be presented because they are subject to the provisions of 14 CFR 191.1.1 Nevertheless, because the critical aspects of the test protocol for the MDNR blind tests were consistent with those used for certification testing of the CTX-5000 SP, the Pd can be compared in a general sense. The blind test results that are available for the MDNR are shown in Table 3-2. The CTX-5000 has passed FAA certification testing and, therefore, meets the FAA's detection requirements. Although the MDNR has not been submitted for certification testing, the blind test results suggest that the Pd for all categories and configurations of explosives required for certification, with the exception of Class A explosives, would be acceptable. The inability to detect Class A explosives is the most significant deficiency of the MDNR. Since the blind tests were conducted, Tensor has refined the detection algorithm on simulated data (Tensor Technology, 1998c); however, in the opinion of the panel, simulation results are not acceptable substitutes for actual test results.

Based on the results of laboratory blind tests of the MDNR (Tensor Technology, 1998b) and certification testing of the CTX-5000 SP (FAA, 1996a), there is no evidence that the MDNR significantly exceeds the detection performance of the CTX-5000 SP for any category of explosives. Furthermore, the cumulative2 Pd for the CTX-5000 SP is higher than for the MDNR. Although some field test results suggest that, in some cases, the detection performance of the CTX-5000 SP combined with an operator may be lower than the performance level of the MDNR, these data cannot be used for comparison because the MDNR has not undergone field tests.

1  

Information subject to the provisions of 14 CFR 191.1 is sensitive but not classified. Information determined by the FAA to be sensitive may not be released without the written permission of the Associate Administrator for Civil Aviation Security (ACS-1), Federal Aviation Administration, Washington, DC 20591.

2  

The cumulative Pd is the Pd of an EDS averaged over all explosives categories.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 26
6 Comparison of Pulsed Fast Neutron Transmission Spectroscopy and FAA-Certified Explosives-Detection Systems In this chapter, PFNTS is compared with the existing FAA-certified EDSs on the basis of their practicality for improving aviation security. All existing FAA-certified EDSs are based on x-ray CT, and all deployed certified EDSs are either InVision CTX-5000 or CTX-5000 SP systems. The data on PFNTS detection capability were provided to the panel by Tensor Technology (testing at the University of Kentucky) and the University of Oregon, and all of the PFNTS operational characteristics evaluated by the panel are based on Tensor Technology's MDNR design. The panel's intention, however, is not to compare Tensor Technology and InVision Technology but to make as comprehensive a comparison as possible between PFNTS and x-ray CT systems in terms of their acceptability for deployment in airports. Performance The motivation for deploying explosives-detection equipment is to improve aviation security. The selection of the equipment is determined by its ability to detect explosives and, at the same time, cause minimal disruptions of airline operations. The FAA's certification requirements reflect both of these requirements (FAA, 1992). Although the specifics of the FAA's certification requirements are classified, the factors that are measured are not. Perhaps the most important factor is the Pd (probability of detection). To pass certification testing, an EDS must be able to detect various explosives configurations at a rate determined by the FAA. Because a high Pfa (false alarm rate) could impede airline operations, the FAA has also set a requirement for a minimal Pfa. Finally, the FAA requires a throughput rate of 450 bags per hour. The detection of explosives is the fundamental performance criterion for comparing competing technologies. The results of certification testing of the InVision CTX-5000 are classified, and some results from blind tests of the Tensor MDNR cannot be presented because they are subject to the provisions of 14 CFR 191.1.1 Nevertheless, because the critical aspects of the test protocol for the MDNR blind tests were consistent with those used for certification testing of the CTX-5000 SP, the Pd can be compared in a general sense. The blind test results that are available for the MDNR are shown in Table 3-2. The CTX-5000 has passed FAA certification testing and, therefore, meets the FAA's detection requirements. Although the MDNR has not been submitted for certification testing, the blind test results suggest that the Pd for all categories and configurations of explosives required for certification, with the exception of Class A explosives, would be acceptable. The inability to detect Class A explosives is the most significant deficiency of the MDNR. Since the blind tests were conducted, Tensor has refined the detection algorithm on simulated data (Tensor Technology, 1998c); however, in the opinion of the panel, simulation results are not acceptable substitutes for actual test results. Based on the results of laboratory blind tests of the MDNR (Tensor Technology, 1998b) and certification testing of the CTX-5000 SP (FAA, 1996a), there is no evidence that the MDNR significantly exceeds the detection performance of the CTX-5000 SP for any category of explosives. Furthermore, the cumulative2 Pd for the CTX-5000 SP is higher than for the MDNR. Although some field test results suggest that, in some cases, the detection performance of the CTX-5000 SP combined with an operator may be lower than the performance level of the MDNR, these data cannot be used for comparison because the MDNR has not undergone field tests. 1   Information subject to the provisions of 14 CFR 191.1 is sensitive but not classified. Information determined by the FAA to be sensitive may not be released without the written permission of the Associate Administrator for Civil Aviation Security (ACS-1), Federal Aviation Administration, Washington, DC 20591. 2   The cumulative Pd is the Pd of an EDS averaged over all explosives categories.

OCR for page 26
A low Pfa is another important requirement for FAA certification. The CTX-5000 SP met the FAA's Pfa requirement during certification testing. The Pfa for the MDNR during blind testing at the University of Kentucky would not have passed certification testing, although the Pfa of the PFNTS in blind tests at the University of Oregon was much lower than the requirement for certification. The bag contents used for the PFNTS blind tests were not the same as the bag contents used during certification testing of the CTX-5000 SP, however, which makes direct comparisons difficult to make. Furthermore, Pfa and Pd are closely correlated and should not be addressed in isolation. The panel believes a credible comparison would require complete receiver-operating characteristic curves for both systems. Unfortunately, this level of characterization is not available for either the CTX-5000 SP or the MDNR. It has been reported that the automated3 Pfa for the CTX-5000 SP is higher in the field than it was during certification testing and that the automated Pfa of the MDNR during blind testing was lower than the automated Pfa for some of the deployed CTX-5000 SP units (DOT, 1998). Alarm resolution is critical to evaluating the performance of a system in an airport. In field tests of the CTX-5000 SP, the issue of alarm resolution was addressed by operator assistance and other airport-specific measures. The issue of alarm resolution for PFNTS has not been adequately addressed in existing studies. However, the Pfa for the CTX-5000 SP-operator combination in field tests was lower than for the MDNR in laboratory tests. Because the spatial resolution of the PFNTS image is low, operator intervention does not lower the Pfa. Therefore, another method of alarm resolution would have to be found. The last major performance metric is throughput rate. The CTX-5000 SP has a throughput rate of 225 bags per hour per machine; in the certified configuration of two CTX-5000 SP instruments, it passes the FAA's requirement for a combined throughput rate of more than 450 bags per hour. Tests of the MDNR concept at the University of Kentucky using a linear accelerator demonstrated a throughput rate of 16 bags per hour. The rate was severely limited by the linear accelerator, which had a low current and, therefore, low neutron fluence. Tensor estimates that with an Ebco TR9D cyclotron accelerator (which operates at a current roughly 30 times higher) the scan time could be as low as eight seconds per bag, which translates to a throughput rate of 450 bags per hour, which is consistent with the FAA requirement. In the panel's judgment, the MDNR could attain the throughput rate required for certification. Operations It could be argued that the most significant operational consideration for deploying explosives-detection equipment is overall cost, which includes purchase price, installation costs, personnel costs, consumables costs, and maintenance costs. Ultimately, every operational characteristic affects cost. For example, if false alarm resolution is a slow, arduous process, then the throughput rate goes down, which could cause flight delays and, therefore, increased costs. Various operational aspects of PFNTS and CT systems are evaluated in the following sections. Costs The first cost incurred by the government or the air carrier is the purchase price of the equipment. As shown in Table 6-1, an InVision CTX-5000 SP costs $1 million; the projected purchase price of the initial MDNR is $2.7 million.4 The installation cost for the CTX-5000 SP ranges from $20,000 for a lobby installation with no modifications to the baggage line to $3 million for a complex installation fully integrated into the baggage-handling system, which requires extensive modifications to the baggage line. The average installation cost for "stand-alone" units was $20,000 to $150,000 per system. Tensor's estimates of the installation cost for the MDNR for the simplest possible airport baggage-line configuration is $323,000 to install the vaulted area to house the MDNR and another $175,000 to modify baggage lines (Tensor Technology, 1998a). Based on the experience of deploying the CTX-5000 SP, the panel is skeptical that the MDNR could be installed for the projected cost, especially if a separate facility is required to house the MDNR. Tensor estimates it will cost $39,000 a year for manpower to operate the MDNR, $15,000 a year for consumable supplies, and $6,000 a year for maintenance (Table 6-1). For deployed CTX-5000s, the manpower costs are $150,000 a year,5 and maintenance costs are $48,000 to $90,000 a year. Reliability and Maintenance Redundancy is a factor that should be considered when comparing CTX-5000 SP installations with a proposed PFNTS system. In many airports, two or more CTX-5000 SP units have been installed, either to increase throughput or to inspect international transfer baggage. The benefits of having multiple CTX-5000 SP units are obvious. If mechanical problems, baggage jams, or other problems render one unit inoperative, the entire airport operation does not have to be shut down. The size and siting constraints of PFNTS systems 3   The automated Pfa is the false alarm rate prior to alarm resolution by an operator. 4   In quantities of 10 or more, Tensor Technology projects that the MDNR could be sold for $1.5 million. 5   This cost includes a full complement of security screeners who may perform other duties when they are not operating the CTX-5000.

OCR for page 26
TABLE 6-1 Baseline Characteristics/Attributes Used in This Assessment Attribute CTX-5000 or CTX-5000 SP Comments Projected Data for MDNR Comments Unit cost ($ million) 1.0   2.7 Initial unit       > 25 Development cost Tensor states $1.8M for assembly and testing of the MDNR prototype. (The additional cost is the panel's estimate of the development cost before airline deployment of a PFNTS production version.)       1.5 In quantities of 10 or more. Installation cost ($ thousand) 20 ± 10 Lobby installation with no modifications. 323 Tensor's estimate for vaulted area.   150 ± 50 Lobby/behind ticket counter installation with modifications. 175 Tensor's estimate to modify baggage lines.   2,000-4,000 Fully integrated installation including cost of one CTX-5000 SP.     Complexity of installation Easy to difficult CTX-5000 series EDSs have been installed in lobbies, behind ticket counters, and in baggage lines. (One baggage-line installation was in a mezzanine [2nd floor].) Difficult Deployment of the MDNR (except on ground level) is probably not feasible because of weight considerations. Compatibility with baggage-handling system Yes Installations in airport terminal lobbies and behind ticket counters are relatively straightforward. Installations that require integrating the CTX-5000 into baggage lines are more difficult. Difficult Placement in baggage-handling station or in separate building. Integration appears to be straightforward, but sharp bends are not compatible with bag movement, while straight paths are not compatible with radiation shielding. Pd(%) Classified   85 Pd for Class B explosives was 94.4; for Class A explosives it was 51.8. Pfa(%) Classified   25 In Tensor blind testing. Operations cost ($ thousand/shift/year), including overhead 48-90 Maintenance policy. 90 Manpower (Tensor states $39,000).   150 Manpower 15 Consumable supplies.   10 Operator training. 6 Maintenance Operators (number/ machine/shift) 1   1 Cyclotron operator. Ancillary manpower 0 Fully integrated installation.       1 Lobby installation and partially integrated installation.       2 Behind counter; one bag handler on front and one on back end.     Educational requirements (years) None No educational requirements, but InVision's recommends that operators speak English and not be colorblind 2 Associate's degree (or equivalent) in a technical field.

OCR for page 26
Attribute CTX-5000 or CTX-5000 SP Comments Projected Data for MDNR Comments On-the-job training requirements 1-2 weeks 1 week for foreign installation. 2 weeks for domestic installations. FAA requirement sometimes stretches up to 4 weeks of training. 6-12 weeks Specialized training. Tensor suggests 6 weeks. Bag throughput (automated bags/hr) 200-250 CTX-5000 SP (111 bags/hour highest throughput, including alarm resolution, observed in the field). > 200 Tensor estimates potential of 1,674 bags/hr.   270 CTX-5500 DS SURE software.       380 CTX-5500 DS CERT software Both softwares certified by FAA.     Down time (%) 2 InVision prefers to describe as"up-time" (98%). < 2 Tensor estimate supported by Ebco Technology (manufacturer of the cyclotron in the proposed Tensor MDNR design) does not include down time of support equipment. Film safe No For ASA > 400 damage to film is evident in prints.     Film tolerant Yes 100, 200, and 400 ASA damage in negatives but not in prints. Appears to be agreeable to vast majority of travelers. unknown Tensor states yes, but more testing required. Alarm resolution Relatively easy CTX-5000 SP provides a cross-sectional image useful for resolving alarms. Moderate to difficult Image produced by stand-alone MDNR is not useful for alarm resolution. Bag could be imaged with x-rays to resolve alarms. System weight (tonnes/tons) 4.3 (4.7)   657 (723) Tensor estimate. System footprint (m2/ft2) 11.3 (125) Machine footprint is 5 m2 (56 ft2) (length = 4.45 m [14.7 ft] and width =1.89 m [6.25 ft]) with an additional 6.2 m2 (69 ft2) required for ramps and console. This does not include space required for baggage handling or the INVIROPAK 93 (1034) Tensor estimate of footprint (length = 11.8 m[39 ft] and width = 7.88 m [26 ft]). Does not include some support equipment requiring about 3.6 m2 (40 ft2). System height (m/ft) 2 (6.7)   3.5 (11.5)   Power requirements 12 kVA; 50-60 Hz; 350-510 V 3 phase   77 kW Tensor estimate. Temperature requirements 10-40 °C (50-104°F)   18-25 °C (64-77 °F) Temperature limits from Ebco literature. Higher temperatures may be permissible. Relative humidity requirements (%) < 80   < 70  

OCR for page 26
Attribute CTX-5000 or CTX-5000 SP Comments Projected Data for MDNR Comments Utilities Electricity and telephone connections.   Not clearly indicated in conceptual design. Clean air, temperature-controlled water, stable power with short cable lengths from supply. Tensor states no problem. Certainly practical but may require additional support equipment located at MDNR. Mean time between failure(hours) 722 Determined during certification testing. Value not obtained from field experience. Unknown Estimate that filament in cyclotron must be changed once per month (which takes approximately 15 minutes). Operational data lacking. Mean time to repair (minutes) 54 Determined during certification testing. Value not obtained from field experience. Unknown   Safety concerns Minimal to none. Meets FDA cabinet x-ray machine specifications. Severe Radiation safety. Shielding is very practical, but safeguards are required to ensure very high radiation levels within cyclotron shielding and that apertures are monitored. Requirements are similar to hospital requirements for cyclotron production of radioisotopes. Radiation monitoring requirements None Meets FDA cabinet x-ray machine specifications. Yes TLDs for monitoring personnel are required. Machine survey is required. Area survey is required. U.S. Nuclear Regulatory Commission license required No   No   Radiation shielding issues None Unit is self-shielded. Important Additional studies required to validate conceptual design. Shielding appears to be feasible with minor changes in design. Public perception of health risks Low Similar to current x-ray systems. Main public concern is that it is not film safe. Moderate Neutron irradiation of bag will generate some protest, but safety can be assured. System does not interface directly with passengers. may prohibit the installation of more than one unit at a single airport. Therefore, if the unit were out of service-for any reason-the baggage-screening operation would be severely compromised. The deployed CTX-5000 SP has 98 percent up time (i.e., 2 percent down time). The mean time between failure (in laboratory testing) for the CTX-5000 is 722 hours, and the mean time to repair is 54 minutes (FAA, 1996a). Tensor estimates that the cyclotron would require ~2 percent down time for maintenance: the filament in the cyclotron would have to be replaced once a month (i.e., mean time between failure is 720 hours), which takes about 15 minutes. The mean time to repair is typically 8 hours for cyclotrons similar to the ones in the Tensor design. Ebco Technologies, the manufacturer of the cyclotron accelerator in the Tensor MDNR design, agrees with Tensor's estimates. The estimates do not address down time associated with the detector array or support equipment. Alarm Resolution The Tensor report does not address the issue of alarm resolution, a very important consideration in the selection and fielding of explosives-detection equipment. The MDNR is at a significant disadvantage in this respect compared with existing image-based detection systems, such as the CTX-5000 SP. The coarse grid image provided by a PFNTS-based system (see Figures 3-1 and 3-2) was of little help to an operator for resolving an alarm during blind tests; thus, the automated Pfa was nearly identical to the Pfa with an

OCR for page 26
operator present. The CTX-5000 is of more help for alarm resolution because it produces a sharper image for the operator to view. The utility of this image for alarm resolution has been confirmed by the significantly lower Pfa of the operator-CTX-5000 combination than the automated Pfa of the CTX-5000 alone (FAA, 1997a). Even if the MDNR' s Pfa is similar to the rates of certified EDSs, the alarm resolution issue strongly suggests that existing CT technology is preferable. To address this issue, the Pfa of the MDNR would have to be much lower than the rate required by the current certification specifications. Even if the MDNR Pfa could be lowered to 1 percent, which is beyond reasonable expectations based on the existing test results, an MDNR in an airport would have to be considered as just one level in a multilevel explosives-detection system. Although additional explosives-detection equipment could be located at a different place along the baggage line, the poor spatial resolution of the MDNR technology would probably require that a higher resolution technique be located in close proximity to the MDNR to resolve alarms. In other words, the MDNR would probably have to be integrated with a high-resolution imaging technology, such as an advanced x-ray or CT x-ray system. The complementary aspects of the x-ray and the PFNTS detection approaches suggest that this combination could be very effective. Airport Integration The location of security equipment on airport property is a primary concern of air carriers and airport operators. Site surveys to identify space for the CTX-5000 SP presented many challenges to the FAA's SEIPT, the airlines, and airport operators. Airport space is at a very high premium, especially at busy airports where security equipment is needed most. Finding the 945 m2 (9,360 ft2) of floor space required for a PFNTS would be much more difficult than finding the 6.7 m2 (67 ft2) required for the CTX-5000 SP. Another constraint would be that only the ground floor could support the 657-tonne (723-ton) weight of an MDNR installation without major floor-support construction. Thus, an MDNR could not be installed in most, if not all, ticket/check-in areas of an airport terminal building, which has been shown to be the most desirable location for screening baggage in certain airports. Another concern is that airlines and airport tenants regularly exchange space and modify terminal configurations. The deployment of a PFNTS-based explosives-detection device would inhibit this flexibility because of the extreme weight and size of the equipment. Baggage Screening The complex process of handling baggage in U.S. airports depends on cooperation between air carriers, including the sharing of baggage-handling systems, which are very labor intensive. In general, the only automated portions of most baggage-handling systems are baggage conveyor belt(s), which run from ticket counters or curbside check-ins to the baggage sorter location, and automated on-demand baggage tag printing. Even "automated" baggage conveyor systems, however, require manual intervention because of baggage jams, mechanical failures, oversized bags, and special articles (e.g., animal kennels). The integration of explosives-detection equipment into the baggage-handling system adds considerable complexity to the operation. The MDNR proposed by Tensor Technology, which produces neutron and gamma radiation, would require that the device be enclosed in a 528-tonne (581-ton) concrete vault with radiation locks and a heavily shielded door (Tensor Technology, 1998a). In the preliminary design, baggage is shown flowing through the unit on an automated belt system with six 90-degree turns inside the vaulted enclosure (see Figure 5-2). A mechanical device would place each bag in the proper (upright) position for scanning. The bag would be stopped, scanned, and then restarted to exit the system. By today's standards, this is a "complex" baggage-handling system. By contrast, the InVision CTX-5000 SP has been successfully installed in 18 U.S. airports in lobby installations, partially integrated installations (e.g., behind the ticket counter), and fully integrated installations (e.g., integrated into the baggage line). Although all of these installations encountered problems, baggage-handling operations were essentially uninhibited by the CTX-5000 SP (NRC, in progress). The performance of existing automated baggage systems—without the added complexity of explosives-detection equipment—clearly shows that bags occasionally get "jammed" together, requiring that the belt be stopped and that manual (human) intervention be used to clear the jam. Experience has shown that the more complex the system is, the more likely jams and mechanical problems are to occur. Therefore, it is likely that baggage jams would occur inside the vault enclosure in the proposed MDNR design, especially because of the changes in direction in the baggage line. Health and safety guidelines would prevent an airline employee from immediately entering the MDNR enclosure to clear a jam. By the same reasoning (and by experience), because the design of the CTX-5000 SP involves no changes in the direction of the bag line inside the CTX-5000 SP structure, baggage jams are unlikely to occur. Furthermore, an operator can easily clear a baggage jam at the entrance or exit of the CTX-5000 SP. Redesigning the baggage-handling component of the MDNR could reduce, but not eliminate, the likelihood of baggage jams inside the MDNR enclosure. Even a redesigned system could not eliminate the delay in getting an employee safely into the enclosure to clear a baggage jam or correct a minor mechanical problem. This delay would have a negative impact on system operating performance, especially on throughput.

OCR for page 26
Cargo Screening The 1996 Valujet accident brought safety and security issues related to cargo shipments on passenger aircraft to national attention (DOT, 1998). Passenger aircraft carry nearly 60 percent of all air cargo, 6 and the air cargo industry in the United States involves linkages between 4,000 air carriers, 3,000 forwarders of air freight, 4,000 repair stations, and 70,000 shippers of dangerous goods (DOT, 1998). Although these issues were reviewed by the White House Commission on Aviation Safety and Security (1997) and by Congress, policies and regulations regarding security screening of cargo and mail are still in a state of flux and will not be reviewed in this report. Based on information submitted to the panel by the FAA, however, the panel assumed that some form of cargo screening for explosives will be required in the future (Fainberg, 1998). As the Tensor report points out, the air cargo problem is difficult to address because of the large variety of sizes and substances that are shipped (Tensor Technology, 1998a). Configurations of cargo, which vary by type of aircraft and cargo shipper, include containerized cargo (e.g., LD-3 containers), palletized cargo, and loose-loaded cargo (sometimes called bulk-loaded cargo). Furthermore, cargo make-up may be done either at the airport or at the cargo shipper's facility, which has implications for the explosives-detection equipment that can be used. Tensor's assessment of the applicability of the MDNR for cargo inspection is based on the assumption that cargo is shipped in LD-3 containers and that each container contains a single class of material (e.g., books or frozen fish). This assumption is based on a rough characterization of air cargo provided by the FAA (1996b), and on this basis the analysis in the Tensor report is valid. However, in actual practice cargo may be shipped in other types of containers, such as the LD-2 container, or as palletized or loose cargo. To complicate matters further, the cargo in a single container may not be uniform. Containers could carry a variety of cargo items. Tensor's assessment, which is based on a few experimental tests of the MDNR with small packages and theoretical extrapolations, may not be valid for all of these variables. In fact, the MDNR has not been tested for full LD-3 containers or other containers of comparable size. Therefore, the capacity of PFNTS to screen cargo for explosives cannot be assessed until additional tests have been conducted. As presently configured, the CTX-5000 SP cannot scan an LD-3 container for explosives because the opening of the CTX-5000 SP is too small to fit an LD-3 container inside. Even if a CT-based explosives-detection system capable of scanning an LD-3 container were developed, it is highly unlikely that it would be effective for detecting explosives because the x-ray attenuation in a container the size of an LD-3 would reduce penetration to the point that an analysis of the contents would not be feasible. When mail, packages, and other cargo are placed on pallets or in containers at the airport, items could be screened by the CTX-5000 SP prior to make-up. Although the CTX-5000 SP may be as effective in detecting explosives in suitcase-sized (or smaller) cargo as it is for passenger baggage, it has not been validated for this use (it should be noted that the panel did not have any data on the effectiveness of the CTX-5000 SP for detecting explosives in cargo). Many of the issues for integrating either the CTX-5000 SP or the MDNR into airports for cargo screening are the same as for passenger bag screening. The restriction on installing the MDNR in lobbies would be less relevant, however, because cargo would be screened elsewhere in the airport. Size, weight, cost, and performance, as well as the impact on airline operations, would all be important considerations in selecting a technology for cargo screening. Based on available data, the panel believes that the deployment of the MDNR or any other PFNTS-based equipment for cargo screening cannot be justified because only a fraction of the total cargo loaded onto an aircraft would be effectively screened. Thus the cost-benefit ratio would be high. The panel also concluded that there is no conclusive evidence that the CTX-5000 SP is an appropriate system for screening cargo. Licensing and Regulations The licensing and operating regulations for x-ray-based EDSs are well understood and are routinely implemented in airports. The federal regulations for the generic operation of a PFNTS-based system with an accelerator and a neutron source are documented in regulatory acts and in the Code of Federal Regulations (CFR): 10 CFR 834, Radiation Protection of the Public and Environment; and 10 CFR 835, Occupational Radiation Protection. Regulations that address the safety of personnel and protection of the environment include the Clear Air Act, the Clean Water Act, the Safe Drinking Water Act, the Radiation Control for Health and Safety Act of 1968, the Occupational Safety and Health Act of 1970, the National Environmental Policy Act, and RCRA. The CFR addresses the implementation details for the production of medical isotopes and gamma irradiators. However, because the radiation-producing system of a PFNTS would not be used for medical purposes and does not have a sealed gamma source, it is not covered by these regulations. Because the use of accelerators for explosives detection is new, no regulatory implementation details have been developed. In the absence of regulations governing the use of neutron-producing accelerators for explosives detection, the panel used the regulations for gamma irradiators as indicators of possible requirements. The scope and potential 6   Air cargo can include the following items: airmail (e.g., sacks, flats, and boxes); express packages; COMAT (company materials); COMAIL (company mail); diplomatic pouches; courier pouches; and baggage (shipped as cargo).

OCR for page 26
impact of some of these regulations are outlined in Box 6-1. If the laboratory-based PFNTS technology were moved into a public area, such as an airport, there would probably be changes in the current regulatory requirements and, perhaps, more specific implementation guidelines, as there are for gamma irradiators used to sterilize materials. In addition to federal regulations on the treatment and control of radioactive materials and sources, some states also have licensing and registration requirements for accelerators. The details of the system design and operation of the MDNR are at the level of conceptual design but do not address regulatory operational requirements and cannot be used as a basis for evaluating the impact of regulations. Until detailed documentation has been prepared on an engineering-level design, the compliance of PFNTS designs with the regulatory requirements must be considered a significant, but not insurmountable, impediment to acceptance by airlines and airports. Regulatory hurdles for PFNTS will begin with the design and construction of the system and extend to the operational BOX 6-1 Selected CFR Regulations Relevant to PFNTS CFR 36.29 specifies that irradiators with automatic product conveyor systems must have a radiation monitor with an auditable alarm to detect loose radioactive sources that are carried toward the product exit. This regulation applies to irradiators that contain sealed gamma-emitting sources. Moderate-energy gammas (< 5 MeV) cannot activate most materials. The radiation monitor for irradiators is focused on early detection of possible leaks from the sealed gamma source. Given the cautionary nature of these regulations (the sealed gamma source material must be doubly encapsulated), a prudent design of a neutron-producing PFNTS system capable of activating material in the vault area will almost certainly incorporate an exit detector to inspect all outgoing bags. An exit monitor would be a prudent measure even if calculations suggest that the normal bag neutron illumination would not reach regulatory limits for activated material because activated room material may have fallen onto the bag. This conservative design is consistent with the design of the thermal neutron analysis system, which also has an exit detector. CFR 36.51 lists detailed requirements for training operators of irradiators. Regulations call for instruction on applicable radiation safety issues, a written test, on-the-job or simulator training, and annual safety reviews. CFR 36.53 requires written procedures for most activities, including the radiation surveys when entering or leaving irradiator radiation rooms, irradiator operations, and facility inspections. CFR 36.55 requires radiation monitoring of personnel with film badges or thermoluminescent dosimeters (TLD) at irradiator facilities. The TLD processor must be accredited by the National Voluntary Laboratory Accreditation Program. TLDs must be processed at least quarterly and film badges at least monthly. CFR 36.57 requires radiation surveys of the area outside the irradiator shielded area at intervals not to exceed three years or whenever the source strength has been changed. This would imply shielding surveys on recently installed cyclotrons and every three years of operation as long as the maximum permissible deuteron beam current is not increased. CFR 36.57 requires portable survey meters at irradiators to be calibrated at least annually. CFR 36.63 requires ''an irradiator operator and at least one other (trained) individual" to be present on site. This suggests that a PFNTS system might require that two trained persons be on site during operations. Cross training of operations personnel could ensure that the second person could perform other duties and not just be on call. A radiation safety officer might also be required to be on call. CFR 36.81 lists detailed requirements for keeping records and requirements for accident reports for irradiator facilities, including requirements for records related to decommissioning.

OCR for page 26
environment. There are no regulations that appear to require an Environmental Impact Statement (EIS) for the MDNR. However, based on experience with the thermal neutron analysis system, completing an EIS for the initial MDNR units would be prudent. Because public perception will be an important consideration in the decision to deploy a PFNTS system, the EIS would assure the public that the system has been designed and will be operated safely and in compliance with all applicable regulations. Some states have regulatory requirements for radiation-producing equipment (e.g., accelerators) that must also be addressed. Although the panel has no reason to believe that an MDNR would be incompatible with the regulations in any state, the differences among the regulations could be an impediment to the acceptance of the MDNR at an airport. Once an engineering-level design becomes available and implementation details have been formulated to ensure compliance with federal regulations, state regulatory agencies would have to be consulted to determine if design changes would be required for compliance with the full range of state regulations. Although none of the anticipated federal or state regulatory requirements is clearly incompatible with the efficient airport implementation of PFNTS technology, if establishing the infrastructure to ensure compliance with regulatory requirements were difficult and costly it could discourage airline and airport authorities from selecting PFNTS explosives-detection technology (even if the system met the FAA certification criteria).