INTRODUCTION

The Ohio DOT (ODOT) has done extensive work on their transportation modeling processes in order to comply with the air quality analysis requirements of the CAA and the recently enacted transportation conformity requirements. On March 10, 1994, Fred Ducca and John Byun of FHWA Headquarters visited ODOT to discuss issues related to conformity and NOx. Chuck Gebhardt represented ODOT. The following are findings from the visit:

1. The ODOT has done extensive work to expand the traditional 4-step transportation modeling process, both in terms of the individual link details and the time periods considered. They have also been extremely thorough in collecting field data to support these model refinements. Traffic volumes and speeds were estimated on an hourly basis. Using this model set, all the Ohio nonattainment areas evaluated showed small increases in NOx for the build compared to the no-build alternative.
2. Based on NOx speed data developed by the California Air Resources Board, ODOT developed a freeway analysis method which increases NOx emission factors associated with ramps/weaving operations, but decreases NOx emission factors associated with mainline operations (see Attachment 2). This method consistently reduced the difference in NOx estimates between build and no-build alternatives (see Attachment 3). The methodology was preliminarily discussed with EPA but until EPA can verify this methodology and modify the MOBILE5 emission factors for all States, the conformity regulations will not permit them to be used.
3. For Toledo, ODOT tested several TCMs to evaluate their ability to reduce NOx. Even though some of the strategies were aggressive (see Attachment 4), none were capable of reducing NOx emissions by 2 percent, even under an assumed reduction in total area auto work trips of 10 percent.
4. The FHWA review team noted that the post processor used by ODOT in estimating freeway speeds (the speed vs. volume/capacity ratio relationship) is similar to the 1985 HCM method (see Attachment 5). The large speed variation based on capacity is responsible for some of the increase in NOx when highway improvements are made.

   However, updates of these speed/capacity relationships are currently underway by the TRB Highway Capacity and Quality of Service Committee. New updates of the freeway curves were approved by the Committee in 1992 and were printed for the Committee on February 7, 1994. The latest research indicates that speed is almost constant with all Levels Of Service until volume reaches the critical level (see Attachments 6 and 7). Also, the Committee adopted increased freeway lane capacities from 2000 passenger cars per hour per lane (PCPHPL) to 2200 PCPHPL for 4-lane freeways and 2300 PCPHPL for 6-lane freeways. Publication of the new material as a formal part of the HCM is expected later this year.

   It was expected that incorporating these updates in the model would reduce the difference in NOx emissions between build and no-build analysis. Also, the entire NOx analysis would need to be re-run within the modeling framework because the assignment process would redistribute traffic among arterials and freeways based on the newly adjusted link speeds. The results of making this is change in Ohio (see Attachment 8) raised the overall NOx estimates for both the build and the no-build cases slightly, but the build alternative became better than the no-build alternative for NOx in Springfield and Toledo, and NOx differences were reduced in the other areas.

5. The FHWA team also noted that hourly NOx emissions on freeways during off-peak periods were relatively high even though overall traffic volume on freeways was low. This effect occurred because of the large percentage of heavy-duty diesel trucks on freeways during evening offpeak periods between midnight and 4 o'clock in the morning (heavy-duty diesel vehicles emit disproportionate amounts of NOx--approximately 40-50 percent of total NOx from highway vehicles). Because speeds during these times were fairly high and NOx emission rates increase rapidly above 80 KPH (50 mph), it was expected that a speed enforcement program would significantly reduce NOx projections.

   The ODOT re-ran the NOx emissions model with revised speed curves for Cincinnati and modeled a strict late night speed limit enforcement. The results are shown below:

   • Total NOx for·580

     build alternative: 99.026 metric tons/day
     no-build alternative: 98.657 metric tons/day
     ---------------------------------------------------------------
     difference: 0.369 metric tons/day
     percent difference: 0.37 percent

      

   • Impact of freeway speed enforcement 88 KPH (55 mph)
     11p.m. - 12a.m. -0.326 metric tons/day
     12a.m. - 1a.m. -0.201 metric tons/day
     1a.m. - 2a.m. -0.113 metric tons/day
     2a.m. - 3a.m. -0.180 metric tons/day
     3a.m. - 4a.m. -0.153 metric tons/day
     4a.m. - 5a.m. -0.153 metric tons/day
     5a.m. - 6a.m. -0.191 metric tons/day

Therefore, speed enforcement for any 3-hour period between 11 p.m. and 6 a.m. would produce NOx reductions greater than the build/no-build difference in Cincinnati.

CONCLUSIONS

1. Updating transportation models to current speed/capacity relationships will lessen the modeled NOx increase associated with the build condition, but not necessarily make it go away. Also, it may generate slightly higher mobile source NOx emission estimates for both build and no-build alternatives.
2. The Ohio NOx model analysis demonstrated that enforcing the 88 KPH (55 mph) speed limit on freeways (where the speed limit is already 88 KPH) between 11 p.m. and 1 a.m. could eliminate NOx problems for the city of Cincinnati. However, caution should be exercised before using this strategy. The program would need to be included as a mitigation strategy that is clearly linked to the build option, and would not otherwise occur. The State DOTs/MPOs would need to coordinate this TCM with EPA's regional office, State and city police departments, and FHWA's regional office to assure that such a program would be acceptable and that all parties agree on the scope and effectiveness of such a program based on public acceptability, limitations on budget, technical difficulties, or legal problems.
3. It is becoming increasingly clear that the analyses required as part of the conformity finding for transportation TIPs and Plans are showing exceedingly small differences in travel and emission estimates between build and no-build alternatives. Refinements to travel models will increase their ability to reflect small differences between options, but will not consistently eliminate the potential for modeled NOx increases for the build option over the no-build. Transportation capital investments and most TCMs may be helpful, but often produce only minor changes in mobile source emission projections, unless the proposals alter travel choices in fundamental ways and affect large segments of the traveling public, or are targeted effectively to vehicles which emit disproportionately large amounts of NOx.

1. ODOT increased emissions associated with ramps and decreased emissions associated with smooth running, (Note: EPA is evaluating this technique.)
2. The ramp speeds are assumed as one half the merge or diverge speed with maximum speed being 92.8 KPH (58 mph) and minimum being 17.6 KPH (11 mph).
3. To better estimate the effect of acceleration or deceleration, adjustment factors are multiplied by MOBILE5A emission factors.

   	Factors for Pollutant
   	HC	CO	NOx
   For Ramps:	1.5	1.5	110
   For Surface Arterials:	110	1.0	110

   For freeways operating in a steady state mode with speed equal to or greater than 72 KPH (45 mph):·580

   • For NOx, the factor is 0.80.·580
   • For HC and CO, the factor is 1.0 at 72 KPH (45 mph) and decreases linearly from 1.0 at 72 KPH (45 mph) to 0.8 at 88 KPH (55 mph) and then increases linearly to 1.0 at 104 KPH (65 mph).·580

Source: OHIO DOT. Chuck Gebhardt
* Units are In metric tons and can be converted to English tons by multiplying by 1.1024.
# ODOT developed factors associated with freeway ramp and mainline operations (see Attachment 2).

Source: OHIO DOT, Chuck Gobhardt

*Total daily mobil source NOX in metric tons. MOBILE4.1 was used for the study.

#Individual TCM9 wore evaluated and compared with 1996 no-build base Cass.

1985 Highway Capacity Manual:

Speed-Flow Relationships under Ideal Conditions

Attachmen*. c.

Source: OHIO DOT, Chuck Gobhardt

* Units are In metric tons and can be converted to English tons by multiplying by 1.1024.

April 5, 1994

MEMORANDUM

There is some concern that future highway mobile source fleet emissions of oxides of nitrogen (NOx) will exceed base year 1990 levels even with moderate growth in vehicle miles travelled (VMT). We have examined this issue and provide the following analysis.

Using the latest version of the MOBILE5 model (March 26, 1993), a base scenario was chosen using the following parameters:

• Summer temperatures (72 to 92 degrees fahrenheit)
• National average fleet characteristics
• Industry average fuel characteristics at 8.7 psi
• National average hot/cold start VMT fractions

Other parameters were varied to investigate their effect on the trend in emissions. Primary in these was the assumed I/M program description, since I/M can affect current and future NOx emission levels. The following I/M program descriptions were used:

Basic

• 1983 program start year
• 40% stringency factor
• 1968 and newer model year vehicle coverage
• No waivers
• 100% compliance rate
• All gasoline vehicle classes covered
• Test-only, biennial inspections
• Idle test procedure (all model years)
• Full anti-tampering program (all components)

IM240 Program

• Same as Basic Program except:
• IM240 test procedure for all model years
• Cutpoints: 0.8/20/'2.0 g/mi HC/CO/NOx

All scenarios were done at 19.6 miles per hour. Non-I/M cases were done at 27 and 50 miles per hour to investigate the potential effect of speed on the Nox results. Also, one case 4as done assuming introduction of new vehicles certified to the ne,4 Low Emitting Vehicle (LEV) standards proposed by California. The model was evaluated every other calendar year from 1990 through 2020. Growth rates from zero to 6% were assumed and applied linearly to the 1990 base NOx levels. The results of the analyses are presented in the attached tables.

Table I shows the non-1/m case at 19.6 miles per hour. In this case a growth rate of 2% will cause NOx emission levels to exceed 1990 base Nox emission levels, but not until calendar year 2020. A 3% growth will cause NOx emission levels to exceed the 1990 base NOx emission levels immediately. Fleet turnover, however, keeps NOx levels close to the 1990 levels until 2010, when the growth in VMT overcomes fleet turnover and emission increase continuously.

Table 2 shows the Basic I/M case at 19.6 miles per hour. The Basic I/M program design reduces NOx emissions by deterrence of tampering behaviour and repairs of tampering with emission control@ devices that control Nox emissions. In this case, as in the nonI/M case, a growth rate of 2% will cause NOx emission levels to exceed 1990 base NOx emission levels, but not until calendar year 2020. A 3% growth will not cause NOx emission levels to exceed the 1990 base NOx emission levels until calendar year 2000. A 4% growth causes NOx emissions to increase continuously. In this case, if it is assumed that in the 1990 base year there was no I/'M program, the 1990 NOx emission target would be 3.000 g/mi. Therefore, if the I/M program were applied after the base year as a control strategy, at a 3% growth, the I/M program would delay the exceedance of the 1990 base levels until calendar year 2010.

Table 3 shows the IM240 I/M case at 19.6 miles per hour. The IM240 I/M program design identifies high Nox emitting vehicles using an IM240 test and requires their repair in addition to identifying vehicles with tampering. In this case, a growth rate of 2% will not cause NOx emission levels to exceed 1990 base NC:-, emission levels until sometime after calendar year 2020 (the ii M4@ of the model). A 3% growth will not cause Nox emission levels t:) exceed the 1990 base NOx emission levels until calendar year 2020. A 4% growth causes NOx emissions to exceed 1990 levels in calendar year 2012. A 5% growth causes NOx emissions increase continuously. As before for the Basic I/M case, if it is assumed that 4@n the '990 base year t.@ere was no 7@/'M program, the emission-target would be 3.000 g/mi. Tlierefore, iO tire I/'M pr-,zram were applied after the base year as a control strategy, at a 3% growth, the I/M przgram would not exceed the 1990 base levels unt4.1 afl-.i.ar calendar year 2020. '@"he exceedance for a 4% growth woul-4@ !me delayed until calendar year 2014. Even a 5* growth would not cause an ex@.eedance unt4.1 calendar year 2008.

Table 4 repeat3 the 'LM240 I/M case at 19.6 miles per h@u" assuming introduction of new vehicles certified to the new Lc w Emitting vehicle (LEV) standards proposed by California. 7hese vehicles will be subject to a moze stringent IM240 exhaust emissions cutpoints resulting in emission rates which will, ^.n average, meet the emission standards for these vehicles at miles. The LEV program is phased in starting in 1994 and is fully operational by 2003. in addit4.on to the NOx reducing effects @f the I/M program, the lower new iehicle NOx standards continues @-he effect of fleet vzehicle turnover. :n this case, a growth rate o.' 5% will cause NOx emission levels to exceed 1990 base NOx emission levels until 2000 when the reduction in emissions due to the LEV program outweighs the VMT growth. The LEV program continues to cause reductions until sometime after calendar year 2020 (the limit of the model). Similarlyt a 6% growth will cause Nox emission levels to exceed the 1990 base Nox emission levels unt4.1 calendar year 2000. But, the LEV program causes a reduction for the period 2000 through 2012. As before for the Basic I/M case, if it is assumed that in the 1990 base year there was no I/M program, the 1990 Nox emission target would be 3.000 g/mi. Therefore, if the I/M program were applied after the base year as a control strategy, up to a 6% growth, the I/M program would not exceed the 1990 base levels until after calendar year 2020.

Most urban areas have fleet average trip speeds greater than 19.6 miles per hour. For comparison, the non-I/M case was repeated assuming an average trip speed of 27 miles per hour and are shown in Table 5. In this case, although the absolute NOx emission rates have changed, the effect of growth on exceedance of the 1990 base NOx emission levels is similar. A similar table done with a speed of 50 miles per hour shoW3 a similar outcome. This demonstrates that the effect of speed on absolute NOx emission'levels is not a major factor in the exceedance of 1990 ba3e NOx emission levels.

Tabld 6 shows the factors used to increase the emission rate3 to reflect increases in VMT as a result of growth. Growth was assumed to be a linear increase in @-MT from the base year level.

cc: T. Newell
C. Radwan
J. Armstrong, ECSB

NOXCAP2.XLS 3/21/94

NOXCAP2.XLS 3/21/94

NOXCAP2.XLS 3/21/94

NOXCAP2.XLS 4/1/94

All Vehicle Feet NOx Emission Rate (g/mi) with. Growth
Without I/M Case (27 mph)

NOXCAP2.XLS 3/21 @'94

All Vehicie Fleet NOx Emission Rate (g/mi) wft Growth

Without I/M Case (50 mph)

Calendar Growth Rate

Year Year 0% 1% 2% 3% 4% 5% 6%

NOXCAP2.XLS 3,'24,'94

NOx Emission Rate

Assumed Linear Growth Factors

Calendar Growth Rate

Year Year 0% 1% 2% 3% 4% 5% 6%