products that are considered dangerous, having hazards analogous to oil but of a different nature, but which are not required to be carried in tanks separated from the hull. The other two categories are cargoes that are required to be carried in tanks separate from the hull because of the greater hazards they present to the safety of the crew, the vessel and the populace in the surrounding area. These types of cargoes include water reactive cargoes, cargoes that become more corrosive in the presence of water, cargoes with high human toxicity, cargoes with unusual flammability characteristics and other properties that make their potential release a serious safety risk. Tank separation from the hull does not necessarily mean double hull protection although that is a likely way of meeting the requirement. The United States has been successful in having this philosophy adopted internationally in the Bulk Chemical Code developed under the auspices of the Inter-Governmental Maritime Consultative Organization.

Products in all four categories may pose potential pollution threats; however, the polluting nature of the products with respect to the marine environment has not been specifically considered. In the future, the Inter-Governmental Maritime Consultative Organization will be revising the criteria for categorization of these cargoes with pollution of the marine environment in mind. In addition, the Environmental Protection Agency's list of hazardous polluting substances may cause the Coast Guard to reevaluate the products on the basis of their pollution potential. At this time, however, the requirement for tanks to be located certain distances from the hull is done for direct safety purposes only.

For the above reasons the Coast Guard is not requiring defensive spaces solely to provide environmental protection in event of grounding or collision accident. As the final environmental impact statement pointed out, for new tank vesels of 70,000 DWT and above segregated ballast tanks are required to climinate the routine event of contaminating ballast water by the addition of that water to unclean tanks. The proper distribution of these segregated ballast spaces to achieve a secondary defensive space benefit in case of accident is a logical extension of the regulations.

I would like to turn now to the other specific items you mentioned as suggestions for improvement of the regulations.

Flue gas inerting systems will be required on new U.S. tankers of 100,000 DWT or more whose keel is laid after 1 January 1975. The feasibility and need of extending this requirement to existing tankers and establishing lower tonnage limits are presently the subject of an inhouse study.

There are no plans at this time to require the installation of lateral thrusters. They are of limited usefulness except at the slowest of speeds when tugs would already be alongside able to apply the same forces. To complement the tugs' forces by adding thrusters may accomplish nothing except to complicate the picture as far as the pilot or master are concerned since it adds another force vector to be resolved by the human computer. Experience has shown that the human will adjust after a period of their use, but the cost-effectiveness of these devices is certainly questionable considering tug availability and the relatively brief time periods during which they would be used.

It is the intention of the Coast Guard to require loran-C receivers when the system coverage is available. The network of stations is expected to be operational in early 1977.

Requiring collision avoidance systems coupled to a radar to relieve deck officers of the problem-solving burden is still under active consideration by the Coast Guard. It must be understood that the devices do not themselves avoid collisions or groundings. Depending upon the degree of sophistication of the device, they can provide alarming services, automatic acquisition of targets, automatic plotting, solutions, data display (both graphical and/or digital),and trial evasion maneuvers. Just which of these features needs to be required and in what format for best assimilation by the human operators remains to be determined.

Please be assured that the Coast Guard is concerned about the protection of the marine environment as an important component of our overall safety responsibilities and that we remain committed to the goal of worldwide climination of operational pollution and minimization of accidental pollution.

Sincerely,

O. W. SILER,

Admiral, U.S. Coast Guard, Commandant.

[APPENDIX DJ

AN AUTOMATIC VESSEL TRAFFIC CONTROL SYSTEM UTILIZING LORAN-C

(Author: Martin C. Day)

ABSTRACT

The increased concern for safety and the ecology of our world waterways creates a need for positive control of shipping traffic in the coastal and inland waterways. This paper describes a Loran-C oriented system to obtain the vessel's present position and the techniques used to utilize this data in a traffic control system. A successful feasibility demonstration of these techniques was conducted in the Fall of 1972 under the sponsorship of the St. Lawrence Seaway Authority of Canada.

INTRODUCTION

Early in 1972 the St. Lawrence Seaway Authority issued the requirement for an Automatic Vessel Location and Indentification System. The system was given the acronym AVLIS.

The system continuously monitors and plots the location and identification of all commercial vessels transiting the St. Lawrence Seaway. The control area originated at St. Lambert Lock, Montreal and terminated at Long Point on Lake Erie, a distance of 425 miles.

The system was required to locate the vessels within ± 250 feet of their true location 95% of the time within the Seaway channels and within ± 500 feet of their true location 95% of the time on Lake Ontario and Lake Erie.

In the Fall of 1972 a feasibility demonstration of a Loran-C oriented AVLIS system was sponsored by the St. Lawrence Seaway Authority to obtain operational verification of the accuracy of Loran C in the Seaway and conclusively demonstrate the feasibility of the AVLIS system concept.

SYSTEM CONCEPT

Preliminary investigations indicated that a system based upon utilization of the existing Loran C chain would be the most advantageous since the Seaway was in a prime Loran C coverage area.

The Seaway was divided into an eastern and western section, operating on an independent basis. The eastern section originated at Montreal Harbour with a Control Center located in St. Lambert. The western section, which includes Lake Ontario, the Welland Canal and a portion of Lake Erie has it's Control Center located in St. Catherine.

As originally conceived, the system consists of the following basic components as shown in Figure 1: Shipboard unit, remote site equipment, and control center equipment.

The Shipboard Unit would receive the Loran C signals, compute the time difference numbers and transmit them along with the ship's identification code over a VHF radio link to a remote site upon a request from the Control Center. The Remote Site Equipment would receive the data and perform the necessary functions required for transmission of the data over leased telephone lines to the Control Center.

The Control Center Equipment contains the circuitry required to receive and check the validity of the data and process the data for use on a visual display. The Control Center equipment would provide a means of interrogating the vessels at pre-set intervals or at a random selected rate. A means to select a particular vessel for interrogation would be provided and the received data from each vessel would then be used to determine it's particular location in the Seaway.

SYSTEM IMPLEMENTATION

The system used for the feasibility demonstration consists of the following equipment: LCR-301 Loran C receiver, shipboard interface unit, VHF transceiver for shipboard installation, VHF transceiver for remote installation, remote interface unit, and control center unit.

A fully automatic LCR-301 Loran C provided the time difference numbers and alarm status bits in BCD form through an accessory connector, received and processed the East Coast chain Loran stations. The stations used were: MASTER-Cape Fear, North Carolina; Secondary A—Nantucket, Massauchusetts and Secondary B-Dana, Indiana. A standard marine 10-foot fiberglas whip was used as an antenna with the LCR-301 coupler.

The Shipboard Interface unit, was connected to the accessory connector on the LCR-301 and performed all the functions of a send/receive modem between the output of the LCR-301 and the VHF transceiver. The vessel's preclearance number was inserted using the front panel thumbwheel switch. Provisions were made to provide outputs to an Hp 5050B digital recorder to record the time difference numbers at a 2 second, 1 minute or 8.5 minute rate. A manual operate switch activated a data transmission to transmit position data from specific check points to evaluate the system positional accuracy. The data was transmitted in groups of 8 bit words using FSK modulation at a 75 band rate.

Data request decisions and automatic keying of the transceiver were also accomplished by the Interface unit. The 4 digit data request received from the Control Center was checked and compared to the number stored in the front panel switch. If the numbers compared, the transceiver was keyed and the position data was transmitted.

Identical 30 watt VHF FM transceivers were used for both the shipboard and remote installations. The remote site transceiver provided the VHF data link between the vessel in the Seaway and the Remote installation. During initial checkout of the system, the transceivers were used for voice communication.

The Remote Interface Unit provides the audio interface between the remote transceiver and the telephone lines. A carrier detector and keying circuit activates the transceiver for transmission of data requests to the vessels.

The Control Center Display Unit provides the means of selecting the vessel to be interrogated, rate of interrogation and displays the received position data and receiver status alarms. Selection of a vessel was accomplished by inserting the assigned pre-clearance number into a 4-digit thumbwheel switch. The controller had the option of selecting an automatic or manual mode of interrogation. In the automatic mode, interrogation was initiated at 45 second intervals. The three numeric displays updated on each data reception to display receiver status, time-difference A, time difference B and the pre-clearance number. Receiver status alarms were displayed to alert the controller of conditions existing on the shipboard receiver that would effect the validity of the position data.

These alarms consisted of lost signal, envelope error and blink alarms. Internal circuitry provided the interface to the telephone lines. The Control Center Unit performed a parity check and bit count on the data as it was received. If invalid data was detected another data request was automatically initiated. A hp 5050B provided a permanent record of all received data. The vessel position was accomplished by plotting the received time difference numbers on a Loran grid overlay on the test area map.

FEASIBILITY DEMONSTRATION

The feasibility demonstrations were conducted in each of the Seaway regions. The Eastern region contained the South Shore canal from St. Lambert to Lake St. Louis and the channel of Lake St. Louis to Lock #4 in Beauharnois. The Remote equipment was located in Beauharnois approximately 23 miles from the Control Center Equipment in St. Lambert. The Western region tests were conducted in the Welland Canal from the Lake Ontario entrance to the control lock at the Lake Erie end. The Control Center Equipment was installed in the existing Control Center in St. Catherines. The prime objectives of the test were:

1. Prove sufficient Loran coverage was available in the Seaway region.

2. Determine if the accuracy requirements could be achieved.

3. Determine the effects of operation in the Seaway locks.

4. Establish the reliability of data transmission over the VHF link within the terrain of the Seaway.

5. Establish the effectiveness of the interrogation technique.

Time difference overlays for the region were generated based on the theoretical propagation paths over land since adequate large scale Loran C charts were not available. The theoretical grids were then displaced based on data gathered by a Loran monitor receiver to correct for differences between actual and theoretical propagation paths. Sufficient data was collected to determine

the shift required to bring the theoretical grid into agreement with the real world. Accurate time differences were thn obtaind along the entire test

area.

Initial system checkout was conducted with the shipboard equipment installed in a van. The equipment was then installed in the Seaway Tug, STORMONT for operational tests. Figure 5 illustrates the system configura

tion.

The Montreal test area originated at the St. Lambert Lock continued through the South Shore canal, across Lake St. Louis and terminated at Lock #4 in Beauharnois. This track covered a distance of approximately 23 nautical miles.

The vessel was interrogated at 45 second intervals by the Control Center. These data requests were sent over a 4-wire phone line from St. Lambert to Beauharnois, a distance of apparently 30 miles. From Beauharnois the data was sent over the VHF link to the STORMOUNT.

Figure 6 illustrates the vessels track through the South Shore canal at a speed of approximately 7 knots. The tracks were obtained by plotting the instantaneous position data as it was received at the Control Center. Examination of the test data shows that there is no apparent error in position due to vessel maneuvers. However, deviations due to noise and bridges were observed. The deviations from the true track of the vessel are totally random with no apparent lags during starts, stops and turns. This was verified by