Formal training had not been provided employees on the operating rules since the SHRTL merged with the GCRTA. Safety Board investigators found a wide difference of opinion among officers and employees as to what rules, if any, were in effect on the Shaker Heights Line. Officers of the operating department of the GCRTA did not express a clear understanding of the rules. Some of the employees believed that some of the old SHRTL rules were still in effect.

Employees were notified of changes in operating rules and procedures by the posting of bulletins at the crew change terminals. Employees were not required to acknowledge receipt of this information or that they understood the changes. Although management and employees generally believed that notices on the bulletin board superseded any rule with which they would conflict, no rule stated this. At the time of the accident, there were no specific rules in effect that governed a singletrack operation. Notices of track work that required all trains to use a single track at a specific location were sometimes posted on the bulletin boards, but they were not required. Notices concerning the single-track operation on July 8, 1977, were not issued.

There were no speed restrictions in effect to govern the operation of trains running against the current of traffic. There was a wide variance in interpretation by employees as to what speed was permitted under this circumstance.

Trains operating on the Shaker Heights Line are equipped with 4channel, two-way radios, which normally are assigned channel 2 for operational control communications. However, at 9:45 a.m. on the day of the accident, radio control at CUT transmitted a directive via channel 2 that all trains monitoring channel 2 switch to channel 1. Westbound Block 12 received this message and made the change. Eastbound Block 4, however, was located under the CUT building at 9:45 a.m., and the motorman stated that he did not receive the transmission. Therefore, his radio remained tuned to channel 2. GCRTA rules do not require that all units acknowledge receipt of such transmissions.

Meteorological Information

The weather on the day of the accident was clear and bright; visibility was good; the temperature was 80° F; the rails were dry.

Survival Aspects

The majority of the 74 persons involved in the accident were riding on the westbound car. Only a few of these persons were not injured to some extent. Of the 60 persons who reported injuries, 5 sustained fractured legs, 2 fractured arms, 2 fractured ribs, 2 fractured noses, and 1 a skull fracture. The other injuries consisted of lacerations and contusions

to the head and neck areas of 23 persons, to the arms and torso areas of 14 persons, and to the leg areas of 11 other persons. Two of the passengers who sustained fractured legs were trapped in the wreckage of the collapsed front end of the westbound car and considerable time was required to remove them from the wreckage. Many of the other persons who sustained lacerations and contusions were treated and released from the hospital the same day.

Tests and Research

Block

A postaccident inspection and test of the block signal system revealed that more than 150 rail bond wires 2/ were broken on the eastbound track within the limits of the single-track operation. signal No. 48 (see figure 1) displayed a red or stop indication regardless of block occupancy, and signal No. 42 displayed a permanent yellow aspect. Block signal No. 54 did not display any aspect; one witness testified that it had been in that condition for about 6 months.

A postaccident test and examination of the braking system components of each car indicated that the brake systems should have operated as intended at the time of the accident. Because of the severity of the cars' structural damage, however, a complete test of each system could not be performed. Testimony of witnesses and inspection of skidmarks at the site disclosed no evidence of malfunction.

Skidmarks on the rails indicated that the brakes were applied on Block 12 and Block 4, 100 feet and 130 feet, respectively, from the collision point. A dense growth of vegetation from 10 to 12 feet high extended 140 feet east of the point of the accident along the inside of the 6° curve. The density of the growth precluded a motorman from seeing across or around the curve as he approached from either the east or the west. The 4 1/2-foot-high concrete wall on the bridge did not interfere with the operators' vision around the 6° curve.

Stopping distances and range of visibility tests were conducted with cars similar to those involved in the accident and under similar conditions to those which prevailed at the time of the collision. The test westbound car approached the point of collision on the eastbound track about 30 mph. This was the estimated speed of Block 12 when its motorman first sighted Block 4. An emergency brake application was made. at the point where the skidmarks of Block 12 began; it required 172 feet for the test car to stop. The stopping point was 52 feet west of the point of collision.

The test eastbound car approached the point of collision on the eastbound track about 30 mph, the estimated speed of Block 4 when its motorman sighted Block 12. The brakes were applied in emergency at the

2/ A metallic connection attached to adjacent rails to insure electrical conductivity.

point where skidmarks of Block 4 were found; it required 212 feet for the test car to stop. The point at which that car stopped was 92 feet east of the collision point. The stopping tests were made with only a motorman and an observer aboard each test car and did not allow for the weight of passengers on either car.

The tests also revealed that when a car of the same type as those in the accident was 160 feet east of the point of collision, it was not visible to the motorman of a train approaching the curve from the west on the eastbound track because of the trackside vegetation. (See figure 4.)

Visibility checks conducted after the vegetation had been removed from the inside of the curve provided sight distances far in excess of the required stopping distances. (See figure 5.)

Other Information

The Safety Board investigated an accident which occurred on the GCRTA on August 18, 1976. 3/ As a result of this investigation, recommendations were made to the GCRTA on August 19, 1977, concerning operating rules and procedures, training, and a method to enforce the rules. In addition, recommendations were made relative to operating trains in occupied signal blocks, the posting of general orders and bulletins, and test facilities for automatic train control equipment. (See appendix A.)

GCRTA's response to these recommendations indicated that they had been complied with or that compliance would be within 60 days. (See appendix B.) The GCRTA is not subject to any Federal or State regulatory authority.

ANALYSIS

However,

The GCRTA did not have operating rules in effect to govern a singletrack operation on the Shaker Heights Line at the time of the accident. The operating department personnel therefore developed their own procedures for the operation. These procedures included the use of two supervisors to notify trains at the crossovers of the need to change tracks. a supervisor was not at Pennsy Crossing when the single-track operation started and was not able to warn Block 4 of the approach of Block 12 on the eastbound track. Neither of the supervisors was specifically instructed to coordinate the operation, and trains were allowed to pass the crossovers at Shaker Square and Pennsy Crossing before anyone had determined that the track was clear. There was not even a method of communications between the crossovers.

3/ "Railroad Accident Report

Rear End Collision of Two Greater Cleveland Regional Transit Authority Trains, Cleveland, Ohio, August 18, 1976" (NTSB-RAR-77-5).

Figure 4. The operator of eastbound Block 4 had this view of the 6° curve approaching the point of collision on the arch bridge (center). There is a transit car 180 feet beyond the collision point, near the truss bridge (background), which is completely hidden by trackside vegetation.

Eastbound Block 4 had no restrictions imposed on its movement, and its motorman did not violate any rules by proceeding on schedule past Pennsy Crossing. The motorman of Block 4 had no way of knowing that he had missed a transmission directing him to change radio channels. Radio control should not assume that all units will receive a transmis

sion that is only broadcast once or twice. If all trains were required to acknowledge such an order, radio control would know when all units concerned had received the message. Apparently, this was not done. There is no evidence that anyone attempted to contact Block 4 after they all failed to contact Block 1. If Block 4 had switched to channel 1, the motorman might have heard the attempt made to stop Block 1 at Pennsy Crossing. Knowing he was approaching Pennsy Crossing, the motorman might have stopped Block 4 west of the crossing and radioed for instructions. He may have been informed of the single-track operation and the accident could have been averted.

The supervisor riding Block 12 should have known that Block 1 already had passed Shaker Square. The supervisors used poor judgment when they allowed Block 12 to leave Shaker Square and operate on the eastbound track against the current of traffic without assurance that all eastbound traffic was stopped at Pennsy Crossing and that the block was clear. The supervisor on board Block 12 exercised poor judgment when he allowed Block 12 to continue west after Block 1 could not be located. Without a "headway" card, the Pennsy Crossing supervisor had to rely on his memory for the schedules of trains; he obviously failed to account for Block 4.

Advance and proper planning is essential for an operation of this nature if it is to be carried out safely. An arrangement such as the one used on July 8, 1977, is especially hazardous when there is no signal protection in one direction, and when there are no safety appurtenances on the car that operate in conjunction with block conditions.

The motorman of Block 4 did not recall the aspect of signal 38 at Pennsy Crossing. Based on the sequence of events and relative positions of the trains before the accident, signal 38 was probably displaying a yellow aspect. This should have alerted the motorman to the fact that the second block ahead was either occupied or that there was a broken rail. However, because he had not been informed that a single-track operation had begun, the motorman would have assumed that a train in the second signal block ahead was moving in the same direction as his train, and that he could proceed to the next signal.

Although the discrepancies found in the signal system near the accident site were not a contributing factor to this accident, the Safety Board concludes that the system requires extensive upgrading to prevent other accidents. The discrepancies indicate poor maintenance practices by the GCRTA. If signals are allowed to deteriorate to the point where they are no longer dependable, operators will ignore them.