Transportation Researh Reord 901 l Impat of Using reeway Shoulders as Travel Lanes on uel Consumption WILLIM R. MCSLN general proedure, based on data derived from the merian ssoiation of State Highway and Transportation Offiials "Redhook", to ompute the savings in fuel onsumption that result from the use of low-ost onversions of urban freeway shoulders to travel lanes is presented. The data re qui red for the analysis are traffi volumes, speeds, and vehile lassifiations before and after the improvements. Three example projets are presented to illustrate the range of benefits in terms of total travel time and fuel onsumption. The projets were implemented in Houston with three different objetives: relieve a main-lane freeway bottlenek, bypass a main-lane queue, and provide priority operation for high-oupany vehiles ( HOVs). The results of the study indiate that major improvements in traffi operations an be ahieved from the additional apaity provided by the shoulder lanes. The magnitude of the improvements depends on the type of use of the lane, the geometri design, and the traffi onditions in the freeway setion to be modified. In the three examples, the annual savings in fuel ranged from 187 87 gal for relieving a bottlenek to 7890 gal for bypassing main-lane queues to 3423 gal for HOV priority opera tions. The savings are the result of improvements in the average speeds of vehiles that use the modified setions. If the improvements an be related to a modal shift for the HOV priority operations, the savings are muh greater. Over the years, design standards have established a typial ross setion for urban freeways with six or more lanes to be 12-ft lanes with - to 10-ft shoulders on both sides of the roadways. However, rising traffi demands and delining finanial resoures have fored transportation agenies to onsider modifiation of these design standards to inrease apaity. One approah that is inexpensive and fast to implement is the onversion of the roadway shoulders to travel lanes. The advantages of inreasing apaity of a setion "overnight" are obvious: inreased servie volumes, lower travel times, and redued traffi delays. There are also disadvantages, both to the users and the transportation ageny: the quality of ride may be less than desirable, the redution of spae available for emergeny parking inreases the potential for aidents that involve disabled vehiles, and the pavement struture of the shoulder may not be designed to handle the inreased loads, thereby resulting in inreases in maintenane osts and a need for early replaement of the shoulder. There was onern that the total aident rate would inrease, but that has not happened 11) There are ways to redue or eliminate these disadvantages: 1. The shoulder an be strengthened and the riding surfae an be improved prior to the onversion, 2. temporary shoulder or speial vehile turnout bays an be onstruted for disabled vehiles, and 3. The traffi loads applied to the shoulder an be redued by restriting the use les and/or restriting the use only, when the additional apaity by heavier vehito peak periods is needed. These measures an also extend the life of the shoulder for many years. There are several reasons for onsidering the use of the shoulder for travel l]j: 1. Overloading the outside lane: The shoulder an be onverted to an auxiliary lane. 2. ypassing main-lanes queues: Queues formed at freeway-to-freeway interhanges an blok traffi movements that are not required to pass through the bottlenek. 3. Relieving freeway bottleneks: reeway setions that have traffi demands that exeed the roadway apaity an benefit from the added apaity. 4. Reduing merge onflits: reeway interhange merge areas that are overloaded or are high aident loations an be improved by using the shoulder on one roadway to move the merge area away from the interhange or to eliminate it entirely.. Providing for priority operation of high-oupany vehiles (HOVs) : shoulder an be onverted to an HOV-only lane to provide a higher level of servie during peak periods. 6. Improving apaities in work zones: Shoulders an be used to restore some of the apaity lost by lane losures for maintenane and onstrution ativities. NITS O USING SHOULRS OR TRVL The primary benefit in the onversion of the shoulder to a travel lane is the inrease in apaity at very low osts. The amount of benefits will vary greatly, depending on the reason for the onversion, the geometri design of the setion, and the traffi onditions. The following examples are presented to illustrate the range of benefits that an be obtained. xample : Relieving a reeway ott.lenek The Southwest reeway (US-9) in Houston was restriped to add an additional lane for a distane of l mile (~) (see igure 1). The setion was four lanes for one-third of a mile and three lanes for two-thirds of a mile. The added lane hanged the ross setion to five lanes for one-third of a mile, four lanes for one-third of a mile, and three lanes for one-third of a mile. In this example, the additional lane had a very high use beause of the high volumes destined for the two exit ramps. The results were an inrease of 22 perent in the servie volumes to 1700 vehiles/h, a redution in total travel time during the 2-h peak period of 1000 vehile-h, and an inrease in the average speeds from to mph over a distane of 3 miles (2 miles upstream of the modified setion). Many projets have reported similar results [Table l!lll, suh as 1. enver, on I-2, with travel time savings of 00 vehile-h, and the speeds inreased from 26 to 46 mph, and 2. Los ngeles, on the Santa Monia reeway (I- 10), with travel time savings of 80 vehile-h, and the speeds inreased from 22 to mph Ill xample : ypassing Main-Lane Queues The West Loop reeway (I-610) in Houston was restriped to provide a bypass lane 0. mile in advane of the interhange of I-10 (see igure 2). The traffi in the evening peak period destined for westbound (outbound) I-10 formed queues of a maximum
2 Transportation Researh Reord 901 igure 1. Use of shoulder lane on Southwest reeway (US-9) in Houston. - --- -U S 9 Southbound ottlenek - - -l - - --- ~---- ---~ Table 1. Summary of operational experiene. Capaity Inreases Inrease ( vehiles/h) elay Redution City reeway efore fter (vehile-h/day) efore fter enver Houston Los ngeles Nashville Portland San raniso I-2 6376 7146 00 26 46 Southwest (US-9) 800 7100 1000 Santa Monia (I-10) 7680 90 80 22 San ernardina 6900 7() 80 Pomona Ventura Golden State Santa na l-6, I-26 I-, I- anfield C-280-7700 800 b 7700 800 410 0 41. 4 7090 100 70 6 8 37 62 0 d - d 20 33 38 ~Inreased apaity by providing a truk Umbing lane. Inreased volumes. ~V i:!hl les inreased by 2 perent and persons inreased by 1 O perent. ri flow. igure 2. Use of shoulder lane on West Loop reeway in Houston. - - - - - l- - - - - Use shoulder as lane to bypass queue... ' I-10 astbound3 length of 1 mile. '!'he traffi destined for eastbound I-10 was delayed unneessarily. Providinq the bypass lane durinq the peak period saved 16 vehi le-h for 8 vehiles that used the 0. -mile bypass. This represents an im.:re<hle in speeds from Lu -!3 rnpi1. xample C: OV Priority Lane The 3 miles of the inside shoulder of the North reeway ( I-4) in Houston was onverted to an HOV priority lane for vehiles that were authorized to use a ontraflow lane (!l (see igure 3). There are 62 buses and 218 vanpool vehiles that use the shoulder lane during the morning peak period. Travel time savings to HOVs were 3.17 min/vehile, whih resulted in a daily savings of 14. 8 vehile-h. In terms of persons arried, the savings were muh more signifiant, resulting in 190 person-h/day. The average sp<!e<.ls fur HOVs im.aease<.l from 26 Lo 48 mph, wil.i.it:= i...i1t:::: ~~t::::t::::uo.i.u i...i1t:= ma.i.u :i..aut:=o Wt=.Lt:= u11c11c:1119"c~. Many other projets an be ited for all of the various reasons for onverting shoulders to travel, and eah would have a unique set of benefits in terms of travel time saved and improvements in average speeds. IMPCT ON UL CONSUMPTION stimates of savings of fuel as a result of the
Transportation Researh Reord 901 3 igure 3. Use of median shoulder as HOV lane on North reeway (1-4) in Houston. -- I- 4 Southbound Median Shoulder as HOV Lane - - - - - t- - - - - - - o-:.;- - o- ~- o- - - -- = = efore 3 Redued Lane Width;- fter Table 2. uel-onsumption rates for vehile type 1 on freeways by LOS and average speed. Table 4. uel-onsumption rates for vehile type 3 on freeways by LOS and average speed. JO 1 2 3 4 0 6 0.0438 0.0468 0.0494 0.067 0.0427 0.044 0.0489 0.019 0.0426 0.0443 0.0471 0.012 0.04 0.0429 0.0 0.0486 0.0433 0.0428 0.0444 0.046 0.3970 0.1649 0.1028 0.0772 0.0641 0.074 0.0431 10 1 2 3 4 0 6 0.1778 0.1928 0.17 0.28 0.18 0.41 0.2128 0.129 0.1676 0.191 0.2167 0.103 0.166 0.1722 0.26 0.167 0.12 0.1646 0.174 3.1346 1.00 0.6 0.3784 0.2963 0.244 Note: Table is based on the proportion of fuel ost to total ost at various speeds as reported in the 1977 SHTO Redhook (~_) and applied to total osts as reported in uffington and MarJand (7). The table is adjusted for 1980 osts of fuel. The fuel osts of the latter report were originally based on the fuel-onsumption rates reported by Claffey (~)and Winfrey (2.). Note: Table js based on fuej-onsu mption rates and fuel osts as a proportion of total osts as reported in thl' J 977 SHTO Red book () and on total osts reported by uffington and Marland (z_). The table is adjllsted for 1980 osts of fuel. Table 3. uel-onsumption rates for vehile types 2 and 4 on freeways by LOS and average speed. 10 1 2 3 4 0 6 0.1486 0.1782 0.1981 0.1412 0.142 0.1687 0.1862 0.1329 0.147 0.13 0.1799 0.120 0.1346 0.1486 0.164 0.1139 0.1281 0.139 0.128 0.676 0.3491 0.2249 0.1772 0.163 0.177 0.1319 Note: Table is based on fuel-onsumption rates and fuel osts as a proportion of total osts as reported in the 1977 SHTO Red book (S) and on total osts reported by uffington and M arland (1_). The table is adjusted for 1980 osts of fuel. improved traffi operations an be determined from the vehile miles of travel at the average speed for the be fore and after onditions. uel-onsumption rates for freeways have been developed from data reported in the 1977 merian ssoiation of State Highway and Transportation Offiials (SHTO) "Redhook" (.2_) and updated to 1980 osts by uffington and Marland (. _) (see table below and Tables 2-4): Vehile Type No. l 2 3 4 Vehile Type esription utomobiles, pikups, and panel truks (2-axle, 4-tire) Single-unit truks (other than 2-axle, 4-tire) Truk trator-semitrailer or trailer ombinations uses The rates in Tables 2-4 provide onservative l'leasures, sine the average fuel-onsumption rates have ontinued to deline with newer vehile models. The fuel-onsumption tables provide a range of values at speeds greater than mph based on the quality of flow as measured by the level of servie (LOS) ny type of freeway improvement an be analyzed if before-and-after data on traffi volumes, speeds, and omposition of traffi are known. or the study of the onversion of shoulders to travel lanes, LOS is used to desribe the after onditions beause of the redution of lateral learanes, quality of roadway surfae, and lane widths. or the example projets, the impat on fuel onsumption is alulated in the following manner. xample : Rel ieving a reeway ottlenek 13 800 vehiles during a 2-h peak period, 2 0-mph before speed (LOS l, -mph after speed (LOS ), 3-mile length of freeway affeted by modifiation, 97 perent type l vehiles (passenger ars, light truks), 2 perent type 2 and 4 vehiles (medium truks and buses), and l perent type 3 vehiles (heavy truks) The hange in the vehile fuel-onsumption rate is as follows (from Tables 2-4): (0.97) (0. 0772-0.0429) + (0.02) (0. 1772-0.1346) + (0.01) (0.3784-0.166) = 0.0363 gal/vehile mile. The average daily savings an be alulated for the total vehile miles traveled during the peak period:
4 Transportation Researh Reord 901 Table. Summary of example projets. Projet esignation No. of Vehiles ffeted Change in Length of Change in Change in uel Projet Travel Time Consumption efore fter (miles) (vhil-h) (gal/year) 13 800 8 280 C' 2 76 26 48 3.0-1000 -187 87 4 0. -16-7 890 48 3.0-14.8-3 423 26 3.0 +4 +11 13 Table 6. -fuel relation for passenger ars. Travel Time uel Consumption uel onomy (min/mile) (gal/mile) (miles/gal).0 1 0.7822 1.278.0 3 0.2849 3.10 12.0 0.184.394 8.6 7 0.1428 7.004 6.7 9 0.1191 8.397. 11 0.10 9.614 4.6 13 0.0936 10.686 4.0 1 0.089 11.637 3. 17 0.0801 12.487 3.2 19 0.07 13.21 2.9 21 0.0717 13.942 2.6 23 0.0686 14.70 2.4 2 0.06 1.142 2.2 27 0.0638 1.667 2.1 29 0.0619 16.149 1.9 31 0.03 16.94 1.8 33 0.088 17.00 1.7 3 0.07 17.387 0.0363 gal/vehile mile x 13 800 vehiles x 3 miles = 103 gal of fuel saved eah weekday. These benefits are only ahieved when the roadway operates without inident. The number of inidentfree days is assumed to be 12/year, whih results in an annual savings of 187 87 gal of fuel. xample : ypassing Main-Lane Queues 8 vehiles use the 0.-mile bypass lane, -mph before speed (LOS ) on inident-free days, 10-mph before speed (LOS ) on inident days, 4-mph after speed (LOS ), and Only type 1 vehiles an use the bypass lane. The average hange in the vehile fuel-onsumption rate for noninident days is as follows: (0.0772-0.0) = 0.0322 gal/vehile mile. or inidents days it is (0.1649-0.0) = 0.1199 gal/vehile mile. The annual savings in fuel onsumption are (0.0322 gal/vehile mile + 0.1199 gal/vehile mile) 8 vehiles) (0. mile) (12 days) = 7890 gal. _...,~~ -- ---... ------, -- -- 'C'... "9'<... 1...,-. un1t n- ~..-..-~ f....,. r..,.,...,.. 62 buses (type 4) and 218 vanpools (type 1) use the 3-mile priority lane eah weekday, 26-mph before speed (LOS ), and 48-mph after speed (LOS ). The before speed is assumed to be the same for inident days. The average hange in the vehile fuel-onsumption rate is (62/280) (0.177-0.187) + (218/280) (0.0628-0.0472) = 0.0163 gal/vehile mile. The annual savings in fuel onsumption are 0.0163 gal/vehile mile x 280 vehiles x 3 miles x 20 days = 3423 gal. xample C': limination of HOV Priority Lane and Return to Passenger Vehiles If the 39 persons who use the buses and vanpools in the HOV lane were to swith bak to passenger vehiles and use the main lanes at an average speed of 26 mph, the fuel onsumption would be 39 persons + 1.3 persons/vehile x 0.0628 gal/ vehile x 3 miles x 20 days/year = 1 232 gal, ompared with 1 097 gal used in the priority lane. Table projets. summarizes the results of the example MINTNNC N RCONSTRUCTION O SHOULR US Routine maintenane of a shoulder lane an be onduted during off-peak periods when volumes an be aommodated on the main lanes of the freeway without additional delay and operating osts to the motorists. However, if the shoulder must be rebuilt, the apaity of the freeway will be less than the original apaity. The additional energy expended during onstrution by traffi that is displaed by the redued apaity an then be estimated. Consider example. To reonstrut the shoulder lane, the apaity of the three original lanes would be redued to 100 vehiles/h during the onstrution perioda This would result in the diversion of 30 vehiles to an alternate route. or this analysis, assume that 1. The freeway would revert to the original operating onditions of mph. 2. The 30 vehiles would use arterial streets with an average speed of 10 mph over a 4-mile trip. The fuel-onsumption rate for this traffi is taken from a reent study of arterial street operation [Table 6 ( 10) l. 3. ll truks and buses will stay on the freeway. T~"= ".! =..i!.~ 1 f'..!~!. C'C"~~ -~!!!'['ti0~ r'!'i0!' t0 t-h~ m0ni-;r:-~tion of the shoulder lane is alulated as follows: (0.0772 gal/vehile mile x 13 386 vehiles + 0.1772 gal/vehile mile x 276 vehiles+ 0.3784 gal/ vehile mile x 138 vehiles) (3 miles) = 33 gal/ day. The daily fuel onsumption during the reonstrution of the lane is alulated as follows:
Transportation Researh Reord 901 (0.1108 gal/vehile mile) (30 vehiles) (4 miles) + (0.0772 gal/vehile mile x 10 386 vehiles + 0.1772 gal/vehile mile x 276 vehiles+ 0.3784 gal/vehile mile x 138 vehiles) (3 miles) = 44 gal/day. Therefore, the reonstrution of the shoulder would ost 901 gal/day. In examples and C, the traffi that uses the shoulder is returned to the main lanes of travel during reonstrution. The only deterioration in operation would be a redution in apaity. or these examples, the bottlenek is further downstream, and thus the freeway operating harateristis in the area would be unaffeted. SUMMRY The onversion of a freeway shoulder to a travel lane is an immediate and low-ost solution for inreasing apaity. The results are higher travel speed, lower total travel times, and redued fuel onsumption. Problems of shoulder pavement deterioration an be lessened by limiting the use of the lane to passenger vehiles during the peak periods only. The impat of the added apaity on fuel onsumption will vary, depending on geometri design, traffi onditions, and use of the shoulder lane. The reonstrution of a shoulder lane after several years of travel may be neessary. However, the daily fuel onsumption during the period of reonstrution should not exeed the amount saved during the use of the shoulder for travel. RRNCS Inrease Capaity. Transportation ngineering Journal, SC, Vol. 106, No. T6, Nov. 1980. 2. W.R. Masland. The Use of reeway Shoulders to Inrease Capaity. Texas Transportation Institute, Texas &M Univ., College Station, Res. Rept. 210-2, Sept. 1978. 3. W.R. Masland and R.G. iggs. reeway Modifiations to Inrease Traffi low. HW, Tehnology Sharing Rept. HW-TS-80-3, Jan. 1980. 4. I-4N Conurrent low Shoulder Lane, Initial indings. Transit Systems evelopment, Metropolitan Transit uthority of Harris County, TX, Projet evelopment Rept. 81-7, ug. 1981.. Manual on User enefit nalysis of Highway and us Transit Improvements. SHTO, Washington, C, 1977. 6. J.L. uffington and G.P. Rith. n onomi and nvironmental nalysis Program Using the Results for the RQ3CP Model. Texas Transportation Institute, Texas &M Univ., College Station, Res. Rept. 210-, Sept. 1981. 7. J.L. uffington and W.. Marland. enefit Cost nalysis: Updated Unit Costs and Proedures. Texas Transportation Institute, Texas &M Univ., College Station, Res. Rept. 2-2, 197. 8. P. Claffey. Running Costs of Motor Vehiles as ffeted by Road esign and Traffi. NCHRP, Rept. 111, 1971, 97 pp. 9. R. Winfrey. onomi nalysis for Highways. International Textbook Company, Sranton, P, 1969. 10. J. Raus. Method for stimating uel Consumption and Vehile missions on Urban rterials and Networks. Offie of Researh and evelopment, HW, Rept. HW-TS-81-210, pril 1981. 1. W.R. Masland. Modifying reeway Geometris to Vehiular uel-consumption Maps and Passenger Vehile leet Projetions LRTO J. SNTIGO The proedures and preliminary results of a study aimed at assessing the fuel onsumption harateristis of passenger vehiles that are representative of the urrent and near-future fleet in order to update the fuel-onsumption models of omputerized traffi simulation and optimization programs are presented. The paper identifies 21 engine-drivetrain ombinations that are representative of 74 perent of the 1979-198 passenger vehile fleet and desribes an in strumentation system that permits the olletion of the mirosopi on the road and laboratory test data neessary to fully assess the real-world fuel onsumption harateristis of vehiles. The problem that this paper disusses is very simple to state: How an we redue fuel onsumption from vehiles operating on a street network? Unfortunately, the answers are quite omplex. Resolving this problem requires a dual approah. irst, we need more energy-effiient vehilesi seond, we need a means of aurately estimating and prediting fuel onsumption from vehiles that operate on a network in order to aurately assess the energy impats of different traffi-ontrol strategies and roadway desiqns. reakthroughs in tehnology ahieved by automotive engineers have provided the means of manufaturing more energy-effiient vehiles. Today automobiles that average 2-3 miles/gal (10.6-14.8 km/l) are ommon (_!) The problems that still remain for the transportation engineer are how to assess and predit vehiular fuel onsumption in any given operating environment and how to enhane roadway designs and traffi-ontrol strategies in order to provide an environment where vehiles an operate more effiiently. The ederal Highway dministration (HW) and others have developed omputer programs that evaluate geometri designs and traffi-ontrol strategies (primarily for urban areas) from environmental and energy onservation standpoints. Use of these models by many users has demonstrated their potential as effetive tools in the development of traffi engineering measures that redue motorist opera ting ostsi fuel onsumptioni osts assoiated with planning, designing, ann implementing new traffi-