6.1 INTRODUCTION

Based on the engineering evaluation of alternate routes, reconstruction of the CSX S line⁹ between Richmond and Raleigh appears to be the most feasible high speed route offering the best alignment between these two cities. This route requires upgrading existing portions of the S line between Richmond and Centralia, and constructing new high speed trackage between Centralia and Raleigh, utilizing the former S line right-of-way. The high speed route would involve the following track and right-of-way segments:

1. Richmond Main Street Station to Centralia -- rebuild existing trackage for higher speeds.
2. Centralia to Norlina -- construct new high speed trackage on former right-of-way.
3. Norlina to Raleigh -- upgrade existing trackage to high speed standards.

In addition to benefiting Amtrak and future state-sponsored passenger services, construction of this high speed passenger route could potentially benefit CSX freight service. If some existing Amtrak service could be shifted from the A line to the S line there may be capacity created for freight operations on the A line. Additionally, there may be opportunities to move selected intermodal freight trains to the rebuilt S line. In order to test these concepts, Wilbur Smith Associates (WSA) performed simulation modeling of passenger and freight service on the two routes. The routes simulated include the A line between Richmond and Savannah via Charleston, the S line between Richmond and Savannah via Raleigh and Columbia, and the connecting Norfolk Southern line between Selma and Raleigh that currently is used by Amtrak.

6.2 STUDY METHOD

This study utilized the Dispatch Planning Model (DPM) software developed by Berkeley Simulation Software, which is widely used by both freight and commuter railroads in the US and Canada. DPM can be used to analyze capacity, to optimize schedules, and to evaluate the impact of track improvements such as higher speed limits or added passing sidings.

The objective of the simulations was to determine how the rail network would perform in accommodating rail traffic. The simulations used alternative scenarios that included construction of the new high speed trackage, shifting of some current passenger and intermodal freight service from the A line to the S line, and the growth in freight traffic anticipated over the next five years.

6.3 ALTERNATIVE OPERATING SCENARIO DESCRIPTIONS

6.3.1 CURRENT A LINE & S LINE OPERATIONS

The CSX A line currently handles most of the freight and passenger traffic for CSX in North Carolina and Virginia. The A line is largely single track, with long double track or passing track segments. The route generally allows passenger speeds of 79 mph and freight speeds of 70 mph for intermodal trains and 60 mph for merchandise trains.

The CSX S line once provided a parallel through route between Richmond and Savannah until trackage between South Collier and Norlina was removed in the mid-1980's. From Richmond to Centralia, the S line is in place, providing a secondary freight route serving industries in the Richmond area. Between Norlina and Raleigh, there is local service to on-line industries. From Raleigh, the S line continues through to Savannah. The S line has more grades and curves than the A line, and has relatively few long sidings capable of accommodating freight trains. Where not limited by curves, passenger speeds reach 79 mph. The bulk of the route sees only one daily passenger train, and primarily local freight service. The exception on this route is between Columbia and Savannah where three loaded and three empty coal unit trains are operated on a typical day.

Amtrak currently reaches the Raleigh-Savannah portion of the S line by operating over Norfolk Southern from Selma, on the A line, to Raleigh. This trackage was included in the simulations. There also is significant freight service on an east-west CSX line through Hamlet that connects with the A line at Pembroke. While the line between Hamlet and Pembroke was not simulated, the freight operations on the line were simulated into and out of the Hamlet Yard, and at the junction at Pembroke where several daily trains join the A line to operate north to Rocky Mount. For the purposes of this study, this will be called Case A.

6.3.2 FUTURE A LINE & S LINE OPERATIONS

Three additional operating scenarios were simulated to represent alternative future cases for the year 2002.

The first (Case B) incorporates growth over the next five years. Two additional intermodal trains in each direction were added to the current schedules with running times similar to current intermodal schedules. The network is the same as base case (Case A) network with no high speed track construction. This case was simulated to provide a basis to measure the impacts of adding the high speed improvements in the remaining two cases.

For the second additional case (Case C), the network was modified to include the proposed high speed line. This case assumed upgraded speed limits between Acca Yard and Main Street Station in Richmond, upgraded trackage and speed limits between Main Street and Centralia on the S line, and new railroad construction from Centralia to Raleigh. Curvature would be eased by new alignments, permitting passenger train speeds of 110 miles per hour or greater. The simulations assumed a maximum of 110 mph over most of the new construction. Where curves limit speeds, the network was coded for operation of high speed passenger equipment with tilting capability, generally allowing speeds 15 to 20 mph faster than standard equipment.

Upgrading and new construction was assumed to correspond to track plans and alignments prepared by TranSystems Engineers and Planners. To improve operating flexibility, WSA assumed three additional sidings in Case C between Raleigh and Petersburg not identified by the TranSystems evaluation. In Case C, only passenger operations would be shifted from the A line to the S line as described below.

1. Four high speed train round trips will operate between New York, Washington, Richmond, Raleigh, and Charlotte, with the frequencies generally 2 to 3 hours apart.
2. Trains 97-98 (Silver Meteor) and trains 52-53 (Auto Train) would move from the A line to the S line to take advantage of the high speed improvements.
3. Trains 91-92 (Silver Star) would remain on the S line using the new trackage north of Raleigh.
4. Trains 79-80 (Carolinian) would continue to operate between Richmond and Charlotte via Rocky Mount and Selma.
5. Trains 89-90 (Silver Palm) would continue to provide service on the A line and would be the only passenger service on the A line between Selma and Savannah.

All passenger trains except Auto Train would use the rebuilt S line through Main Street Station in Richmond.

Finally, a third future case (Case D) included all the high speed improvements north of Raleigh, and assumed extension of three sidings between Hamlet and Columbia to lengths greater than 1.0 mile, to accommodate some added freight service over the S line. In Case D, the same passenger operations would be shifted as in Case C; additionally three intermodal trains in each direction would utilize the S line between Petersburg and Savannah.

Where the rebuilt S line adjoins the A line at Centralia and Petersburg, crossovers between the lines would allow operating flexibility, and permit potential freight service on the S line to bypass Main Street Station and avoid the steeper grade between Main Street and Acca Yard.

The assumptions for each case are summarized in Table 6-1.

Table 6-1: A Line & S Line Capacity Analysis Operating Assumptions

Simulation Case	Passenger Trains	Freight Trains
Case A (Base case)	Current operations	Current operations
Case B (Base network with freight traffic increase)	Current operations	Current operations, plus two additional intermodal round trips
Case C (New high speed trackage, Richmond, VA to Raleigh, NC)	Add four high speed round trips, and shift most current A line service to S line	Current operations, plus two additional intermodal round trips
Case D (New high speed trackage, Richmond to Raleigh, and extend sidings, Hamlet, NC to Columbia, SC)	Add four high speed round trips, and shift most current A line service to S line	Current operations, plus two additional intermodal round trips, with three intermodal round trips shifted to the S line

6.4 MODEL RESULTS

The Dispatch Planning Model was used to simulate train operations for the four scenarios described above.

Under Case B, two intermodal freight trains have been added in each direction. These new trains, together with existing intermodal and passenger trains, result in considerable congestion around Florence, SC during the night.

In Case C, the assumption is that the S line improvements have been completed and the Amtrak Silver Meteor and Auto Train have been shifted to that line, freeing up some A line capacity. The proposed S line passing sidings for the Richmond-Raleigh territory, together with existing passing sidings south of Raleigh, appear to handle meets between passenger trains in a satisfactory manner only if the run on schedule. Considerable delays, however, occur to coal trains in the Savannah-Columbia territory under this scenario.

In Case D, three intermodal trains are also rerouted to operate via the S line. This results in a significant reduction in A line train activity compared to the Case C, providing additional capacity and improvement in the performance of merchandise trains that remain on the line. The rebuilt line between Richmond and Raleigh with 5-mile passing tracks spaced 15-20 miles apart appears to have sufficient capacity. However, the territory south of Raleigh, and particularly south of Hamlet, is extremely congested with delays of up to four hours occurring to expedited intermodal trains due to lack of sidings. The resulting 30-mile spacing of long sidings (the same as presently exists between Columbia and Savannah) is clearly inadequate for the levels of passenger and freight train operations proposed for the S line in this case.

The relocation of Amtrak services from the A to the S line in case C and the further removal of intermodal trains in case D result in lower delays to merchandise trains which remain on the A line. However, these figures clearly illustrate that there will be considerable delay to intermodal and bulk freight operations that would use the S line unless additional places to meet or pass trains are provided in the territory south of Raleigh.

In Cases C and D, passenger service largely would operate over the new high speed trackage north of Raleigh, and on the balance of the S line south of Raleigh. Trains 89-90 (Silver Palm) however, would remain on the A line to serve Charleston, SC and other communities on that route. The running time differences are discussed below by train and are summarized in Table 6-2.

Trains 52-53 (Auto Train) require longer elapsed times running via the S line because of meets with other passenger trains, and because their maximum speed of 70 miles per hour does not take full advantage of the high speed trackage north of Raleigh. Performance would improve if more sidings were available to reduce delay time for meets, but overall time running via the S line is not likely to improve on current running times via the A line.

Table 6-2: A Line & S Line Passenger Train Elapsed Run Times
Train	Route	Cases A & B (Existing track)	Cases C & D (Improved S line)
52 (Auto Train)	South End-North End	9:01	10:21
53 (Auto Train)	North End-South End	8:38	10:29
79 (Carolinian)	Richmond-Raleigh	3:28	3:33
80 (Carolinian)	Raleigh-Richmond	3:17	3:45
89 (Silver Palm)	Richmond-Savannah	8:14	8:02
90 (Silver Palm)	Savannah-Richmond	8:33	8:41
91 (Silver Star)	Richmond-Savannah	11:22	8:28
92 (Silver Star)	Savannah-Richmond	10:26	8:38
97 (Silver Meteor)	Richmond-Savannah	8:32	8:17
98 (Silver Meteor)	Savannah-Richmond	8:41	9:01
02 (SEHSR)	Raleigh-Richmond	----	2:02
06 (SEHSR)	Raleigh-Richmond	----	2:02
09 (SEHSR)	Richmond-Raleigh	----	2:15
10 (SEHSR)	Raleigh-Richmond	----	2:14
11 (SEHSR)	Richmond-Raleigh	----	2:01
12 (SEHSR)	Raleigh-Richmond	----	2:02
13 (SEHSR)	Richmond-Raleigh	----	2:06
15 (SEHSR)	Richmond-Raleigh	----	2:03
Note: Passenger run times are station to station, with Richmond times shown for Staples Mill Road

Trains 79-80 (Carolinian) operate via Rocky Mount and Selma under each case, but running via Main Street Station in Richmond adds a few minutes to their elapsed times.

Trains 89-90 (Silver Palm) continue to operate between Richmond and Savannah via the A line with no significant change in run times. Operation via Main Street adds a few minutes. Elapsed times vary slightly between the cases because of meets in different locations.

Trains 91-92 (Silver Star) benefit significantly from operating over the new high speed trackage north of Raleigh instead of the circuitous route via Selma. These trains can reduce their running times by about 3 hours by shifting to the S line north of Raleigh.

Trains 97-98 (Silver Meteor) have comparable elapsed running times via either route. Using the high speed trackage is balanced by running at generally slower average speeds over the S line south of Raleigh.

The additional high speed service through Virginia and North Carolina would take full advantage of the direct routing via Henderson, and the elapsed times of 2 hours 1 minute to 2 hours 15 minutes between Richmond (Staples Mill Road) and Raleigh are attainable with tilt technology equipment. The difference in time represents delays caused by meets with other passenger trains. Fine tuning of the schedules and the exact locations of the sidings can reduce the longer run times by a few minutes. The simulation shows that elapsed run times between Richmond's Main Street Station and Raleigh, including stops at Petersburg and Henderson, can be as low as 1 hour 43 minutes.

Elapsed operating times for selected intermodal freight trains are shown in Table 6-3. As can be seen, trains which are shifted to the S line in Case D take between two and five hours longer for their run compared to the A line routing. The primary reason for this longer trip time is delays due to meets with passenger trains and other freight trains. In fact, without interference, an intermodal freight will make the run between Acca Yard and the South (Savannah) end of the network in 10 hours 6 minutes.

Table 6-3: Comparison of Elapsed Operating Times, Selected Freight Trains

* Indicates trains operating via S line in Case D.

6.4.1 ADDITIONAL SCENARIOS NOT EVALUATED

The simulation modeling performed under this study only evaluated a limited range of operating options that might be possible. The primary concerns were to simulate operations using the proposed high speed trackage north of Richmond, and to examine the related passenger service options that would be made possible by construction of the high speed line.

6.5 SUMMARY AND RECOMMENDATIONS

Based upon this study, Amtrak, NCDOT and VA DR&PT have the following conclusions and recommendations.

High speed travel between Raleigh and Richmond is possible with improvements to the route. The Berkeley model analysis confirms that rebuilding the S line (along with other right-of-way improvements identified in the TranSystems engineering evaluation) would cut the current travel time from 3:27 to two hours or less.

Some track improvements will be necessary beyond those proposed by the engineering evaluation to accommodate passenger and freight service. Assuming limited growth in CSX freight activity, multiple passing sidings will be necessary along the S line to assure that there are minimal delays for freight traffic diverted from the A line.

Switching some passenger trains to the S line will improve service. Improvements between Raleigh and Richmond will provide a faster and shorter route for the Silver Star. Unless additional capacity is placed on the S line south of Raleigh, however, no other passenger trains on the A line could take advantage of this new route.

Additional capacity beyond current levels will be needed on the S line to assist freight traffic. Preliminary analysis shows that the S line south of Raleigh does not have passing sidings long enough to warrant switching of current A line traffic to the S line. Longer sidings and further schedule analysis will be needed to determine the best freight use of a rebuilt S line.

More analysis is necessary when the CSX acquisition of Conrail is completed. The amount of freight traffic along the eastern seaboard will likely increase with the takeover of Conrail by NS and CSX. Additional analysis will be necessary to determine the effects of this new traffic on the SEHSR system.

⁹For the purposes of this capacity analysis, the CSX S line is the former route of the Seaboard Air Line Railroad from Richmond to Florida via Petersburg, Henderson, Raleigh, Hamlet, Columbia, and Savannah. The CSX A line is the former route of the Atlantic Coast Line Railroad from Richmond to Florida via Petersburg, Rocky Mount, Florence, Charleston, and Savannah. These two railroads merged to form the Seaboard Coast Line Railroad, which eventually became part of today's CSX system.