Frye Island

The ferry is our only way to get on and off Frye Island. What would a day, week, month, or whole season be like on Frye Island without the ferry system? There would be no island access unless you were fortunate enough to have a private boat with docking space on the mainland. The ferry landing and access road would not be available either for docking or parking, because that space would be needed for repairs to the ferry system. The value and enjoyment that we derive from our properties are completely dependent upon our ability to access them, and as things stand now we have only one means of access—the ferry system.

An alternative system for access to Frye Island should be evaluated. The main benefits of an alternative access system would be:

One alternative is a passenger ferry system supported by the appropriate infrastructure (parking lots on the mainland and island, public docks and possibly an island van). We would be able to park our cars on the mainland and still gain access the Island. We would be able to maintain access to our properties if the vehicular ferry system were to become disabled. In addition to providing back-up transport, a passenger ferry system could benefit us in a variety of other ways:

Island growth has put increased demand on the ferry service, especially on week-ends. This trend will only continue. It is prudent for us to compare the costs and benefits of attempting to improve the existing ferry system with the costs and benefits of developing alternative ways to access the island.

The Transportation Committee had its first meeting in July of 2005. After hearing from the town manager Wayne Fournier what was expected of us, we developed the following mission statement:

“The mission of the Frye Island Transportation Committee is to assess the current town infrastructure usage, predict the impact from continued community growth, and recommend strategies to cope with future demand.”

We decided that our first task would be to assess the current utilization of the ferry system. That assessment is the principal subject of this report. This report also details the system improvement options and alternatives that we considered and which of them we recommend.

Island growth, in terms of dwelling units, averaged 10 units per year from 1970 to 2000. (Town of Frye Island Comprehensive Plan, July 2002) In recent years, the Frye Island Growth Ordinance has limited the number of new housing starts to 15 units per year. If we assume the number of dwelling units on Frye Island now is 475, and the maximum number will be 750, then the island should reach its maximum development in 20 to 30 years if growth follows its historical path.

In an attempt to generate a mathematical model of current ferry utilization we used data from different sources:

· Past (pre-2004) data that had been collected for the ferry (ticket counts reported in the Comprehensive Plan);

Run data for the ferry was collected in 2005 by the ferry crew and is available; however it has not been entered onto a computer at the time of this analysis, consequently the 2005 data were not used herein.

It is important to emphasize that the nature of the data collected pre-2004 differs from that collected in 2004, which differs again from that collected since 2004. This heterogeneity in the nature of the available information limited our ability to analyze trends from year to year. It was crucial to the present analysis that the Town invested consistent effort (employees such as managers, mates and ticket takers kept critical and complete records) during 2004 to obtain detailed data on ferry usage from which we could develop some useful conclusions.

After reviewing the data that were available, we found that the run data collected by the mates was the most helpful in creating a picture of the ferry utilization pattern run by run, hour by hour, and day by day. This detailed data collected for the entire 2004 season included a vehicle count by type of vehicle on a run by run basis in both directions (to and from the island). We judged this data to be more helpful than ticket sale data or number of runs, because we could analyze what types of vehicles were using the ferry and when they were using it.

2004 was the first complete season that the mates collected specific vehicle data for each individual run. We recommend that the Town of Frye Island resume collecting ferry run data with this level of detail because such data will assist with:

Historically, decisions concerning ferry operations have often been made based on guesses, intuition, criticisms, and anecdotal information, with no objective data subsequently collected to evaluate the outcomes of those decisions.

Reliable data collected consistently, year over year, is needed to quantify ferry utilization increases or decreases. The only year to year data that exists is ferry revenue figures. However, we recognize that the use of ferry revenues has three major limitations:

We used the 2004 ferry run data to estimate the sustainable unidirectional rates of ferry operation. This was done by identifying times of peak demand when either one or two ferries were in operation and counting the number of runs made in one direction for a particular time frame, then multiplying the number of runs by 9, the maximum number of cars that can be carried by a single ferry.

The data show that one ferry can move as many as 27 cars in one hour in one direction on a sustainable basis hour after hour. Two ferries can move as many as 45 cars in one hour in one direction on a sustainable basis hour after hour. For short intervals, the ferry system has moved up to 54 cars in one hour, but it was able to sustain that rate only for a one hour timeframe on one or two occasions a month. Thus, for practical purposes, the sustainable maximum unidirectional rate of ferry system operation is 45 cars per hour.

For an example consider the ferry runs on Sunday, 8/22/04, from 5 pm to 8 pm from the island (Figure 1 below). For the three hour period shown, the two ferries in operation were able to initiate at least 5 runs per hour in each of the three consecutive hours; at 9 cars per run the result is a sustainable maximum unidirectional rate of ferry system operation is 45 cars per hour. This assumes that the additional loading time that would be required to achieve full loads on the 5 runs that were carrying 6 to 8 cars would not reduce the number of runs in those hours.

The maximum unidirectional rate of ferry system operation was 36 cars per hour in periods when there was substantial bidirectional demand. The reason for this is probably that loading time at both sides reduces the number of initiated runs per unit time. When the demand is unidirectional (i.e., Fridays to the island and Sundays from the island), and the loading time is on one side only, the observed unidirectional rate is 45 cars per hour.

We feel it is important to state that we are reporting the rates of operation that we observed in the data. We are not suggesting that the ferry should operate at any particular rate. Appropriate consideration should be given to the safety of the ferry, its crew, its passengers and operating conditions when choosing a rate of operation, in addition to consideration for the wear and tear on the equipment for any particular operational strategy.

Ferry capacity is the total number of car equivalents that can be moved by the ferry system over a finite measure of time. To calculate various ferry capacities, we used unidirectional and bidirectional rates of ferry operation and different periods of time. For daily ferry capacities, we assume that the ferry will operate for 16 hours per day and for the monthly capacity calculation we assume that there are 30 days per month. The three ferry capacity calculations that are relevant to the rest of our analysis are as follows:

Throughout the rest of our analysis, we will refer to these theoretical maximum ferry capacities as we consider current and predicted future ferry utilization.

We feel it is important to point out that these capacities are theoretical and to achieve these capacities, the operation of the ferry system would have to be flawless and MOST importantly ferry demand would have to be evenly distributed over the respective time frames. Perfect distribution of demand in any transportation system is not possible, thus as ferry demand approaches maximum theoretical daily capacity wait times will be substantial.

In 2004, the ferry system moved 47% of all vehicles for the entire six month season in just the months of July and August. Please see Appendix A to view the total monthly usage by different types of vehicles. Note that in the off peak months of May, June, September, and October, total monthly ferry demand would have to more than triple in any given month before we would exceed the existing ferry capacity. Therefore we chose to focus the detailed daily and hourly analysis of ferry utilization on the peak months of July and August.

Figure 2 below shows a more detailed picture of demand for ferry use during high season—average number of car equivalents using the ferry is given for each day of the week. A “car equivalent” is defined as the space that a single car occupies on the ferry; an oversized vehicle will occupy more than one car equivalent. Thus, the data for the different types of vehicles are analyzed in car equivalents to take into account the fact that certain pieces of heavy equipment require as much space on the ferry as two or three cars, or that captains may decide to limit additional cars from the run. In keeping with everyone’s experience, the peak days of the week for unidirectional ferry transport of car equivalents during high season 2004 were Fridays to the island and Sundays from the island. In 2004, the peak day to the island was Friday July 2, with 474 and from the island was Monday July 5, with 481.

Total average ferry demand on Friday is greater than on Sunday. However Sunday demand is more concentrated because ferry users tend to want to leave the island at around the same time on Sundays, whereas Fridays typically have a more evenly distributed arrival pattern. Also, the ferry system carries construction traffic up to 2 pm on Fridays during high season, whereas on Sundays it carries essentially none. The conclusion from the data is that the ferry system exhibits greater utilization and shorter wait times on Fridays than on Sundays because of traffic distribution patterns.

Interestingly, while there are individual days (for example, Fridays and Mondays of holiday weekends) of extreme demand, on average the demand is spread out over the week in a fairly even manner. See the Average bidirectional utilization (Total Avg in Figure 3) from day to day ranges only from 478 to 598 car equivalents. This may be attributable to a variety of factors such as:

We conclude that, currently, on an average daily basis, during the months of July and August, we use about 50% of the maximum daily bidirectional ferry capacity available when it operates 16 hours a day. Stated another way, if daily average bidirectional utilization levels double from what they are today then we will exceed the theoretical maximum bidirectional daily ferry capacity.

Weighted Daily Ferry Demand
(from Tues July 6, 2004 to Sun Aug 22, 2004)
(weighted into car equivalent units; see App B for Weighting Factors)

Demand to the Island


Day:	Sun	Mon	Tues	Wed	Thurs	Fri	Sat

	128	238	243	243	328	349	295
	136	230	219	258	302	376	277
	177	269	268	224	263	405	309
	141	235	243	261	320	391	328
	127	215	225	244	280	386	313
	140	236	216	227	321	376	280
	152		215	236	305	339

Avg:	143	237	233	242	303	374	300


Demand from the Island

Day:	Sun	Mon	Tues	Wed	Thurs	Fri	Sat

	303	266	322	266	331	209	209
	330	304	229	265	264	196	208
	335	224	258	224	231	234	278
	339	255	253	263	267	264	255
	315	280	220	255	234	222	269
	321	260	220	227	283	250	235
	344		215	247	260	191

Avg:	327	265	245	249	267	224	243

Total (To + From Demand) Avg:	469	502	478	491	570	598	543

The Frye Island Ferry system has historically operated with two primary but conflicting goals:

These two goals are conflicting because the most cost effective way to run the ferry system is with 100% full ferries for every run, but the only way to guarantee 100% full ferries for every run is to always have people waiting in line to use the ferry.

When one sees two empty ferries pass, one can’t help but think the ferry service is being mismanaged and the town is wasting money. One only needs to look at the ferry data and consider the ferry staffing requirements to understand why we may occasionally see two empty ferries pass from time to time. It is not rational to expect otherwise.

One reason for two empty ferries passing is the highly variable nature of ferry demand. The 2004 ferry run data shows that, even at peak times, demand is highly variable and can result in empty ferries passing as was the case for ferry traffic to the island on Friday August 20 from 5:30 pm to 7:45 pm. (shown in Figure 3 below), which shows that the ferry system was running every 15 minutes to the island and was nearly full for every run except for the 6:45 pm and 7 pm runs which were nearly empty. The overall demand during this time frame quite clearly justifies running two ferries, however, if one had been watching the ferry from 6:45 pm to 7 pm, one would have seen basically empty ferries passing (the 6:45 boat from the island carried 1 car and the 7pm boat from the island carried 0 cars) on successive runs.

The second reason for empty ferries passing has to do with ferry staffing requirements and short spikes in demand. There are times when two ferries are needed to handle two or three hours of demand, however when a second ferry is put into service that crew must have a minimum of 4 hours of work. As an example consider ferry traffic from the island on Tuesday August 17, 2004 (Figure 4 below). Between 4 pm and 6 pm on that afternoon, there were 53 cars that traveled from the island to the mainland, however, between 6pm and 8pm there were only 23. The second boat was clearly justified to satisfy the hours between 4pm and 6pm and clearly wasn’t needed from 6pm to 8pm. However, when we put the second boat into action, the crew must have at least a 4 hour shift.

The largest data set that we have to represent long term ferry utilization patterns is the annual revenue figures as reported for the Frye Island ferry system compared with the number of cottages for the respective years. Within the data available to us, there was a stable subset of information from 1994 to 1997, when there was a 4.6% yearly increase in ferry revenues corresponding to a 4.3% yearly increase in total cottages. Considering that construction was steady over this time period and that ferry fees and pricing structure were constant, we conclude that the increased utilization of the ferry over this period of time

Members:	John Riley, Chairman Carl Hommel, Secretary Rose-Marie Boyd Ron Clough Richard Flowerdew Joe Markoski Brian Nisula David Treacy
Contributors:	Wayne Fournier John Crosby

run -	a one way ferry trip, either to or from the Island
ferry system -	a transport mechanism capable of moving people and/or vehicles across water
passenger ferry system -	a transport mechanism capable of moving people and their belongings, but not vehicles, across water
demand -	the number of vehicles that want to move onto or off the island
unidirectional -	use of the ferry system in one direction
bidirectional -	use of the ferry system in both directions
car equivalent -	space on the ferry required to transport one car