5. INTEGRATION Modelling

INTEGRATION traffic modelling software was used to simulate traffic operations in the Primary Study Area for typical weekday a.m. and p.m. peak periods in order to identify sections subject to recurrent congestion. A summary of the methodology and results is presented below. Additional details are provided in the INTEGRATION Modelling Calibration Report, in Appendix E.

5.1 Methodology

INTEGRATION is a trip-based microscopic simulation model, designed to trace individual vehicle movements through a road network from trip origin to destination. The model incorporates two major components, the network and the origin-destination (O-D) matrix, as described in the following sections.

5.1.1 Model Network

The INTEGRATION network is a representation of the physical road network within the primary Study Area. It is coded as a series of roadway links, interconnected by nodes, with operating parameters for each link coded so that the model's simulation will approximate actual operating conditions.

The network coded for this study includes the Highway 417 mainline from west of Highway 416 to west of Anderson Road, as well as all on and off ramps within these limits (not including ramp/arterial road intersections). Link lengths and lane arrangements were scaled from the MTO base plans and the aerial photo orthomosaic.

Key parameters used in coding the network include:

• Maximum Permissible Speed:
  • 120 km/h on Highway 417
  • 120 km/h on the high speed on/off-ramps ¹
  • 80 km/h on the direct ramps¹
  • 60 km/h on the inner ramps¹
  
  
• Speed at Capacity:
  • 80 km/h on Highway 417
  • 80 km/h on the high speed on/off-ramps
  • 60 km/h on the direct ramps
  • 40 km/h on the inner ramps
  
  
• Jam Density:
  • 125 veh/km
    
• Capacity:
  • 2400 vph/lane on the Highway 417
  • 2400 vph/lane on the high speed on/off-ramps
  • 2100 vph/lane on the direct ramps
  • 2000 vph/lane on the inner ramps

The capacity values noted above are the theoretical maximum flows that could be achieved on model links; actual modelled flows will approach but never reach these values. It is necessary to input these theoretical maximums in order for the simulation model to reproduce the highest observed traffic volumes.

The INTEGRATION network created to model existing conditions was used without modifications for modelling of 2011 conditions. For the 2021 horizon year, the only modification incorporated into the network was the addition of the Hunt Club Road interchange. The new interchange was assumed to be located approximately in line with an easterly extension of the existing Hunt Club Road and was assumed to be designed to current MTO geometric design standards.

---

¹ For a majority of ramps, the link coded in the model coincides with the acceleration or deceleration lane. The operating speed on an acceleration or deceleration lane is higher than the operating speed on the corresponding ramp, and is often higher than the ramp design speed.

---

5.1.2 Modelling Periods

Based on a review of the patterns of existing traffic volumes within the Study Area, two three-hour time periods were selected for INTEGRATION modelling. Demand volumes were generated for an a.m. peak period of 6:00-9:00 a.m. and a p.m. peak period of 3:00-6:00 p.m. These time periods capture the build-up to recurring peak period congestion, the critical time period for identification of factors contributing to congestion. It is recognized that congestion may continue, and will ultimately dissipate, beyond the ends of the modelled periods.

The a.m. and p.m. peak hour traffic volume projections for 2011 and 2021 were extrapolated to estimates of three-hour demands for the a.m. and p.m. peak periods, to be used in the INTEGRATION modelling. The parameters used for the extrapolation were derived from the pattern of existing traffic volumes and are summarized in Exhibit 3.8 of the INTEGRATION Modelling Calibration Report (Appendix E).

While the proportions of the volumes occurring during each of the hours in the peak periods can be expected to change over time, there is no firm data available to allow quantification of the changes. Use of the existing volume distribution was considered to be preferable to applying an arbitrary volume distribution. It should be noted that this approach may underestimate the demands in the time periods before and after the peak hour, thus under-representing the effects of peak spreading as demands increase. Some peak spreading is represented in the INTEGRATION model, since vehicle demands that exceed the capacity of the entry link are queued for later entry.

5.1.3 Origin-Destination Matrix

The Origin-Destination (O-D) matrix identifies the traffic demand between each feasible origin-destination pair within the network for each 15-minute interval within the time periods to be simulated. Since detailed O-D data was not available for this study, the required O-D matrices were estimated from observed or projected mainline and ramp volumes using the WatOD software. This software uses an iterative process to assign volumes to each O-D pair until the total link volumes match the observed volumes, within a specified tolerance level (see calibration section, below).

The volumes used for the existing conditions O-D matrix estimation were from the September 2001 ramp counts completed by Pyramid/TSH and the Fall 2000 mainline counts provided by MTO. The recorded hourly volumes were averaged across Tuesday and Wednesday to produce an average weekday hourly volume scenario, representative of volume conditions observed between Monday afternoon and Friday morning. All of the averaged weekday mainline hourly volumes were increased by 0.8% to reflect traffic volume growth between 2000 and 2001.

The 2011 and 2021 O-D matrices were developed using the balanced mainline and ramp volume projections for the a.m. and p.m. peak hours, expanded to three-hour peak period traffic volumes as described in Section 5.1.2. The peak period traffic volume projections were run through WatOD to produce an a.m. and p.m. peak period O-D matrix for each horizon year.

5.1.4 Calibration

Two types of model calibration were carried out on the existing conditions INTEGRATION model:

• Calibration of the modelled volumes to the observed volumes; and
• Calibration of simulated traffic operations to actual traffic operations.

The volume calibration was completed by adjusting the demand volumes in the initial O-D matrices produced by WatOD until the modelled volumes were within ±15% of the observed volumes. The initial O-D matrices produced modelled volumes within ±15% of the observed volumes on all mainline sections and on all but five ramps in the a.m. peak period and six ramps in the p.m. peak period. The volume calibration effort reduced this to three ramps in the a.m. peak period and five ramps in the p.m. peak period. Further improvement was not possible, as congestion limited the volumes able to use the affected ramps.

Calibration of the existing conditions model to observed operating conditions was carried out to ensure that the model accurately reflects areas known to experience recurrent congestion and does not show congestion in areas known to operate well in peak periods. This calibration was based on consultant team observations and input from Ministry staff. It was carried out by making minor adjustments to link parameters, particularly capacity, to better reflect actual link-specific operating conditions. The calibrated model was found to accurately simulate existing peak period operations throughout the Study Area, with the following exceptions:

• Highway 417 eastbound, west of Walkley Road, a.m. peak period -- Operations analysis shows that the congestion in this area is the result of a capacity deficiency at a downstream intersection on Walkley Road (see Section 6.3.1). Since the ramp terminal and downstream intersection were not included in the model network, the simulation does not fully capture this congestion.
  
• Highway 417 westbound, east of Highway 416, p.m. peak period -- The congestion observed in this area appears to result from the operation of the weave between Highway 416 and Moodie Drive. This weaving area is beyond the west limit of the model network, so its effects are not simulated.

The calibrated model was used for the 2011 and 2021 scenarios without further adjustment.

5.2 Results

5.2.1 Level of Service

A level of service (LOS) was assigned to each Highway 417 mainline section for each 15-minute simulation period using the density statistics produced by INTEGRATION and the LOS criteria from the 2000 Edition of the Highway Capacity Manual. (See Section 6.1.3 for a listing of the density ranges associated with each level of service.) The LOS results were reviewed to identify sections that operate at LOS F, congested operation, during one or both peak periods, in each horizon year. The congested sections were classified as experiencing "Congestion" or "Major Congestion" according to the duration of LOS F operations, as follows:

• "Congestion" -- LOS F for 0.5 to 1.5 hours during a three-hour peak period; and
• "Major Congestion" -- LOS F for 1.5 to 3 hours during a three-hour peak period.

The results are summarized in Table 5.1 for the a.m. peak period and Table 5.2 for the p.m. peak period. The results are also summarized on schematic maps in Appendix F. A discussion of the results is provided following each table.

In reviewing the results summarized in the tables and schematic maps, it is apparent that congestion is expected to generally increase throughout the Study Area in future years. However, some locations may be indicated as having improved operating conditions in a later horizon year compared to an earlier one. This condition generally results from a combination of two factors:

• EMME/2 modelling may forecast lower volumes on certain ramps in future years due to shifts in travel patterns in response to network congestion, mode split changes, City road network improvements (2021) and/or demographic changes; and
  
• When a section upstream of a previously identified bottleneck becomes congested in a future year, it limits the amount of traffic that can reach the previous bottleneck during the congested period. This metering effect produces a diminished level of congestion in the downstream section.

These effects occur in only a few sections of the Study Area and are small relative to the overall degradation in traffic operations expected in future years.

Table 5.1 - Summary of a.m. Peak Period Congested Locations
Congestion/Major Congestion*

Eastbound

	Location	2001	2011	2021
1	Hwy 416 off-ramp - Richmond Road	-	Congestion	Congestion
2	Richmond Road - Pinecrest Road / Greenbank Road off-ramp	-	Major Congestion	Major Congestion
3	Pinecrest Road / Greenbank off-ramp - Woodroffe Avenue off-ramp	Congestion	-	-
4	Woodroffe Avenue Interchange	Congestion	-	-
5	Woodroffe Avenue on-ramp - Maitland Avenue	Major Congestion	Congestion	Congestion
6	Maitland Avenue - Carling Avenue off-ramp	Major Congestion	Major Congestion	Major Congestion
7	Upstream of Parkdale Avenue off-ramp	Congestion	Congestion	Congestion
8	Upstream of Rochester Street off-ramp	-	Congestion	-
9	Upstream of Bronson Avenue off-ramp	Congestion	-	-
10	Bronson Avenue - Kent Street	Congestion	Congestion	-
11	Kent Street - O'Connor Street	Major Congestion	Congestion	-
12	Upstream of Nicholas Street off-ramp	Congestion	-	-
13	Downstream of Nicholas Street on-ramp	Congestion	-	-

Congestion/Major Congestion*

Westbound

	Location	2001	2011	2021
14	Anderson Road - Hunt Club Road off-ramp	-	Congestion	Major Congestion
15	Hunt Club off-ramp - Walkley Road off-ramp	-	Congestion	Major Congestion
16	Walkley Road off-ramp - Innes Road on-ramp (Inner)	-	Major Congestion	Congestion
17	Innes Road on-ramp (Inner) - Ottawa Road 174 off-ramp	Congestion	Major Congestion	Major Congestion
18	Ottawa Road 174 Interchange	Congestion	Major Congestion	Major Congestion
19	Ottawa Road 174 on-ramp - St. Laurent Boulevard on-ramp (Inner)	Congestion	Major Congestion	Major Congestion
20	St. Laurent Boulevard on-ramp (Inner) - Vanier Parkway on-ramp (Direct)	Major Congestion	Major Congestion	Major Congestion
21	Vanier Parkway on-ramp (Direct) - Nicholas Street off-ramp	Congestion	Major Congestion	Major Congestion
22	Nicholas Street Interchange	Congestion	-	Congestion
23	Nicholas Street on-ramp - Metcalfe Street off-ramp	Congestion	Congestion	Major Congestion

* Major - LOS "F" observed for duration more than 1.5 hours in a modelling period of 3 hours
Congestion - LOS "F" observed for duration 0.5 - 1.5 hours in a modelling period of 3 hours

The following are some key observations from the a.m. peak period INTEGRATION modelling results:

• Eastbound Highway 417 from the Highway 416 interchange to Pinecrest Road/Greenbank Road is expected to become an area of major congestion by 2011 due to high traffic volumes and weaving conditions between the Richmond Road interchange and the Pinecrest Road/Greenbank Road interchange.
  
• Eastbound Highway 417 from Pinecrest Road/Greenbank Road to Woodroffe Avenue experiences recurring congestion under existing conditions due to high traffic volumes (average 1863 vph/lane during 7-8 a.m. west of Woodroffe Avenue). Based on the 2011 and 2021 simulation results, this portion of highway may not experience congestion in future years due to the major congestion upstream of this location, at the Pinecrest Road/Greenbank Road off-ramp.
  
• Eastbound Highway 417 from Woodroffe Avenue to Carling Avenue experiences recurring congestion under existing conditions due to high traffic volumes (average 2140 vph/lane during 7-9 a.m. west of Maitland Avenue). This section of Highway 417 will continue to experience recurring congestion through 2011 and 2021 due to high projected traffic volumes in the future years (average 2333 vph/lane in 2011, west of Maitland Ave) and considerable weaving upstream of the Carling Avenue off-ramp.
  
• Eastbound Highway 417 from Parkdale Avenue to Kent Street experiences recurring congestion under existing conditions due to high traffic volumes and considerable weaving (average 2070 vph/lane during 7-9 a.m. west of O'Connor Street). This section of Highway 417 will continue to experience recurring congestion through 2011 and 2021, due to high projected traffic volumes (average 2333 vph/lane in 2011, west of Kent Street) and considerable weaving upstream of the Rochester Street off-ramp, Bronson Avenue off-ramp, Kent Street off-ramp, O'Connor Street off-ramp and Nicholas Street off-ramp. Based on the 2021 simulation results, this portion of highway may not experience congestion in the long term due to metering of the flow by the major congestion upstream of this location, at the Pinecrest Road off-ramp and Carling Avenue off-ramp.
  
• Westbound Highway 417 from Anderson Road to Walkley Road will likely begin experiencing recurring congestion by 2011 due to high projected traffic volumes (average 2150 vph/lane in 2011, south of Walkley Road). Once the Hunt Club Road interchange is constructed (2021 scenario), westbound Highway 417 from the Hunt Club Road off-ramp to Walkley Road is likely to develop into an area of major congestion due to high projected traffic volumes (average 2500 vph/lane in 2021, south of Walkley Road) and considerable weaving.
  
• Westbound Highway 417 from Innes Road to Ottawa Road 174 presently experiences recurring congestion due to the merge of Highway 417 and Ottawa Road 174. Once the congestion builds up at the merge, the traffic backs up as far as Innes Road. This section of highway is expected to experience increasing congestion through 2011 and 2021 due to increasing traffic volumes through the Highway 417/Ottawa Road 174 merge.
  
• Westbound Highway 417 from Ottawa Road 174 to Metcalfe Street presently experiences recurring congestion due to high traffic volumes and significant weaving. This section of highway is likely to experience increasing congestion through 2011 and 2021 due to high projected traffic volumes (average 2075 vph/lane in 2021, west of St. Laurent Boulevard) and significant weaving.

Table 5.2 - Summary of p.m. Peak Period Congestion Locations
Congestion/Major Congestion*

Eastbound

	Location	2001	2011	2021
1	Upstream of Hwy 416 off-ramp	-	Congestion	Major Congestion
2	Hwy 416 off-ramp - Richmond Road	-	Congestion	Congestion
3	Upstream of Pinecrest Road off-ramp	-	Major Congestion	Major Congestion
4	Pinecrest Road / Greenbank Road Interchange	-	Congestion	Congestion
5	Pinecrest Road / Greenbank Road on-ramp (Direct) - Woodroffe Avenue off-ramp	-	Congestion	Major Congestion
6	Woodroffe Avenue off-ramp - Maitland Avenue off-ramp	-	Major Congestion	Major Congestion
7	Maitland Avenue off-ramp - Carling Avenue off-ramp	-	Major Congestion	Major Congestion
8	Carling Avenue off-ramp - Parkdale Avenue off-ramp	-	Major Congestion	Major Congestion
9	Parkdale Avenue off-ramp - Upstream of Rochester Street off-ramp	-	Major Congestion	Major Congestion
10	Upstream of Rochester Street off-ramp - Bronson Avenue	Congestion	Major Congestion	Major Congestion
11	Bronson Avenue - Metcalfe Street on-ramp	Major Congestion	Major Congestion	Major Congestion
12	Metcalfe Street on-ramp - Nicholas Street off-ramp	Major Congestion	Major Congestion	Major Congestion
13	Nicholas Street Interchange	Major Congestion	Major Congestion	Major Congestion
14	Downstream of Nicholas Street on-ramp	Major Congestion	Major Congestion	Major Congestion
15	Downstream of Nicholas Street on-ramp - Riverside Drive on-ramp	-	Major Congestion	Major Congestion
16	Riverside Drive on-ramp - St. Laurent Boulevard	Major Congestion	Major Congestion	Major Congestion
17	St. Laurent Boulevard - Ottawa Road 174	Major Congestion	Major Congestion	Major Congestion

Congestion/Major Congestion*

Westbound

	Location	2001	2011	2021
18	Hunt Club off-ramp - Walkley Road off-ramp	-	-	Congestion
19	Walkley Road on-ramp (Inner) - Innes Road	-	Congestion	Congestion
20	Innes Road - Ottawa Road 174 off-ramp	-	Congestion	Major Congestion
21	Ottawa Road 174 Interchange	-	Congestion	Major Congestion
22	Ottawa Road 174 on-ramp - St. Laurent Boulevard on-ramp (Direct)	-	Congestion	Major Congestion
23	St. Laurent Boulevard on-ramp (Direct) - Vanier Parkway off-ramp (Direct)	-	Major Congestion	Major Congestion
24	Vanier Parkway Interchange	-	Major Congestion	Major Congestion
25	Downstream of Vanier Parkway on-ramp (Direct)	Congestion	Major Congestion	Major Congestion
26	Nicholas Street Interchange	-	Congestion	Major Congestion
27	Downstream of Nicholas Street on-ramp	-	Congestion	Major Congestion
28	Downstream of Nicholas Street on-ramp - Metcalfe Street off-ramp	-	Congestion	Major Congestion
29	Metcalfe Street off-ramp - Bronson Avenue	-	Major Congestion	Major Congestion
30	Bronson Avenue - Downstream of Rochester Street on-ramp	Congestion	Major Congestion	Major Congestion
31	Downstream of Rochester Street on-ramp - Upstream of Parkdale Avenue off-ramp	Congestion	Major Congestion	Major Congestion
32	Upstream of Parkdale Avenue off-ramp	Major Congestion	Major Congestion	Major Congestion
33	Parkdale Avenue Interchange	Congestion	Major Congestion	Major Congestion
34	Parkdale Avenue on-ramp - Island Park Drive	Major Congestion	Major Congestion	Major Congestion
35	Island Park Drive - Carling Avenue off-ramp	Congestion	Major Congestion	Major Congestion
36	Carling Avenue off-ramp - Woodroffe Avenue	Major Congestion	Major Congestion	Major Congestion
37	Downstream of Woodroffe Avenue on-ramp (Direct)	-	Congestion	Major Congestion

* Major - LOS "F" observed for duration more than 1.5 hours in a modelling period of 3 hours
Congestion - LOS "F" observed for duration 0.5 - 1.5 hours in a modelling period of 3 hours

The following are some key observations from the p.m. peak period INTEGRATION modelling results:

• Eastbound Highway 417 from Rochester Street to Vanier Parkway presently experiences recurring congestion due to high traffic volumes (average 1725 vph/lane during 3-5 p.m. west of Metcalfe Street). Significant weaving is observed from the Metcalfe Street on-ramp to Vanier Parkway off-ramp due to the close proximity of the ramps. This section of highway will continue to experience congestion through 2011 and 2021 due to higher projected traffic volumes (average 1925 vph/lane in 2021, east of Nicholas Street).
  
• Eastbound Highway 417 from St. Laurent Boulevard to Ottawa Road 174 presently experiences significant weaving due do the close proximity of the on-ramp at St. Laurent Boulevard and the split of Highway 417/Ottawa Road 174. Once the congestion builds up at the split, the traffic backs up as far as Vanier Parkway. Modelling of the 2011 and 2021 horizon years indicates that traffic may back up all the way to Highway 416 in the future.
  
• Westbound Highway 417 from Bronson Avenue to Woodroffe Avenue presently experiences recurring congestion due to high traffic volumes and significant weaving (average 1505 vph/lane during 3-5 p.m. west of Maitland Avenue). This congestion is likely to extend eastward up to Ottawa Road 174 by 2011 and up to Innes Road by 2021 due to high projected traffic volumes and significant weaving (average 1900 vph/lane in 2021, west of Maitland Avenue).
  
• The section of westbound Highway 417 from the Hunt Club Road off-ramp to Ottawa Road 174 is projected to experience congestion throughout by 2021, due to high projected traffic volumes (average 2150 vph/lane at Ottawa Road 174 in 2021).

As noted above, some highway sections that currently experience recurring congestion are projected to have improved traffic operations in future years due to metering of the traffic demand by major upstream congestion. In these cases, capacity improvement of the upstream section can be expected to result in the redevelopment of congestion in the downstream section. INTEGRATION modelling will be used to identify such interactions and ensure that all impacts of proposed improvement alternatives are assessed.

5.2.2 Corridor Performance

INTEGRATION was also used to produce measures of the performance of the entire model network for each horizon year, to provide indications of the expected changes in congestion and impacts on mobility throughout the primary Study Area as traffic volumes grow and their distribution changes. The two measures produced by the model are:

• Average Corridor Speed, calculated as follows: Total vehicle travel time is computed as the summation of trip times across all vehicle trips that are completed within the simulation period. The total vehicle kilometres is computed as the sum of the trip lengths for all vehicle trips that are completed within the simulation period. Average Network Speed is calculated by dividing the total vehicle kilometres by total vehicle travel time.
  
• Average Corridor Stops, calculated as follows: Each time a vehicle decelerates, the drop in speed is recorded as a partial stop. The total vehicle stops are computed as the summation of the number of vehicle stops per trip across all the vehicle trips that were completed within the simulation period. The average vehicle stops are computed as the total vehicle stops divided by the total vehicle trips.

Table 5.3summarizes the corridor performance measures for the modelled a.m. and p.m. peak periods in each horizon year.

The model indicates that increasing congestion and degradation in mobility can be expected over the next 20 years. For example, average network speed between 16:00 and 17:00 is projected to decrease by 26.2 km/h (39%) by 2011 and by 32.1 km/h (47%) by 2021. The impact is less during the a.m. peak period; for example, the average network speed between 08:00 and 09:00 is projected to decrease by 7 km/h (11%) by 2021. While the congestion increases in the a.m. peak period are projected to be less severe than in the p.m. peak period, they are still highly significant when considering the traffic volumes using Highway 417 through the Study Area each morning.

Table 5.3 - INTEGRATION Corridor Performance Statistics

a.m. Peak Period Averages

Time Period	2001 Speed (km/h)	2011 Speed (km/h)	2021 Speed (km/h)	2001 Stops (%)	2011 Stops (%)	2021 Stops (%)
06:00-07:00	105.6	106.2	104.9	4.5	4.2	4.6
07:00-08:00	81.7	78.3	72.6	8.3	8.5	9.0
08:00-09:00	64.4	60.4	57.4	10.5	10.2	10.8

p.m. Peak Period Averages

Time Period	2001 Speed (km/h)	2011 Speed (km/h)	2021 Speed (km/h)	2001 Stops (%)	2011 Stops (%)	2021 Stops (%)
15:00-16:00	80.7	63.4	58.4	10.3	12.2	13.4
16:00-17:00	68.0	41.8	35.9	11.4	14.7	16.2
17:00-18:00	59.2	30.7	27.4	12.2	16.4	17.7

The model indicates a minor improvement in overall corridor performance during the first hour of the a.m. peak period in 2011. This result appears to be caused by a redistribution of traffic volumes, as a result of volume balancing and the travel pattern changes predicted by the EMME/2 model. The redistribution produces a more uniform flow through key sections of the Study Area, delaying the onset of congestion in these areas. The overall improvement in corridor performance is not considered to be significant and is not sustained into the second hour of the peak period or the 2021 horizon year.

---

Contents | Summary | Introduction | Data Collection | Demand Forecasting | Traffic Volumes
INTEGRATION Modelling | Operations Analysis | Safety Review | Conclusions