CHAPTER TWO CONGESTION IN SCOTLAND
2.1 Both Scotland's trunk road network and its urban network are subject to congestion (Scottish Executive, 2006 p7). Such congestion is localised in both time and space. As congestion affects the performance and quality of the transport system through increased travel times; deterioration in the 'driving experience' with stop-start conditions; and reliability problems (leading to travel time variability and large unexpected delays), data sources are required that capture these impacts in order to describe the locations and time periods where congestion occurs. A review of the available literature indicates only one paper (Scottish Executive, 2005) which has attempted to assess the level of congestion in different parts of Scotland using the same objective criteria - and this has only a limited focus: the most congested parts of the trunk road network. Local studies associated with the development of Local Transport Strategies ( LTSs) and Road Traffic Reduction Act ( RTRAs) targets can also report on congestion, as can STAG Part 1 and 2 reporting procedures for proposed schemes whose objectives are to reduce congestion. Such reports have been excluded from this review because different criteria for measuring congestion can be applied in different studies, and therefore there is no objective manner to compare different locations and secondly the scale of the survey that would be required warrants a study in its own right.
2.2 As a result of a lack of studies using the same criteria to measure congestion throughout Scotland, the contribution that other available data sources make towards describing where congestion occurs in Scotland have been considered. Aside from the traffic count data upon which the Executive's trunk road congestion indicator report is based (Scottish Executive, 2005), three further sources have been identified: the Scottish Household Survey ( SHS) ( MORI Scotland et al., 2003-4), the Transport Model for Scotland ( TMfS) (Lumsden, 2005) and journey time data (at a national level) held by and surveyed by ITC Holdings (see http://www.itisholdings.com). The first source gives a measure of delay, whilst the second and third sources can potentially give a measure of journey speed/time and variability in journey speed/time. A data source that considered the manner that congestion impacts on the 'driving experience' has not been identified, nor has it been possible to locate a data source that provides freight specific information. Within the constraints of the current project, additional analysis of the SHS and TMfS data has been undertaken to give background information on traffic delay in Scotland. Previous research experience with the journey speed data held by ITIS Holdings (for example, Grant-Muller, 2005) has been good and as a result this may warrant consideration as a future data source for measuring the journey time and reliability impacts of congestion, particularly if used to give more detailed information on the performance of particular sections of the network.
Perceived delay
2.3 Since 2003, as part of its travel diary the Scottish Household Survey has asked car drivers whether or not they were delayed by traffic congestion on their journey and if so to quantify that delay. The responses to this question reflect perceived delay, as there is no objective measure to the delay perceived beyond that reported. Table 2.1 indicates the proportion of peak hour trips that respondents indicated were delayed by congestion by local authority, whilst Annex 1 presents a more detailed analysis of the 2003 and 2004 SHS datasets. The salient points that can be drawn from Table 2.1 and Annex 1 are set out below.
- 11.5% of trips by road in Scotland experience some form of congestion related delay, whilst 88.5% of trips experience no congestion related delay. The average delay across all trips is 1.3 minutes, however, the average delay for those who actually experience some delay is 11 minutes.
- The delay varies over the different road user groups. Higher proportions of commuters (18%) and business/work related trips (17%) experience delay compared to trips with 'other' trip purposes (8%).
- The delay varies by time period - with trips occurring during weekday and morning peaks experiencing the highest chance of being delayed. On average 25% of trips in the weekday AM and PM peak are delayed compared to only 8% at other times of the day.
- The delay varies in a geographic context. Travellers with a destination in the Glasgow RTP, the Aberdeen RTP and the Edinburgh RTP experience the largest number of delays. Within each of these RTPs the largest number of delays are experienced in the cities themselves - with Aberdeen having the highest proportion of its trips delayed. The geographic variation in the proportion of trips experiencing delay during peak hours is very marked with over 40% of trips with a destination in Aberdeen being delayed in the morning peak, whilst less than 10% of trips in Dumfries and Galloway, Argyll and Bute, the Shetland Islands and the Orkney Islands being delayed.
- Average delay per trip follows the patterns set out above - i.e. average delay per trip is highest in the RTPs related to Aberdeen, Glasgow and Edinburgh. The range of delay is from 2 minutes per trip for trips with a destination in Glasgow and Aberdeen to 0.1 minute per trip for trips in the Shetland Islands.
- Interestingly however for those people who are delayed the average delay is broadly the same across the whole of Scotland with a range of 7.5 to 12.1 minutes. This suggests that certain 'capacity pinch-points' give rise to localised delay in all parts of Scotland. Clearly however the number of people affected and the number of capacity pinch points varies geographically - giving rise to the geographic spread in the proportion of travellers experiencing delay.
Table 2.1 - Proportion of trips delayed by congestion by Local Authority (peak hour trips only)
PEAK HOUR TRIPS ONLY
Council area of destination | Whether part of car/van trip delayed due to traffic congestion |
|---|
Yes | No |
|---|
Aberdeen City | 42.2% | 57.8% |
|---|
Edinburgh, City of | 38.3% | 61.7% |
|---|
East Renfrewshire | 33.3% | 66.7% |
|---|
Glasgow City | 33.2% | 66.8% |
|---|
Midlothian | 32.7% | 67.3% |
|---|
Falkirk | 31.1% | 68.9% |
|---|
Renfrewshire | 30.9% | 69.1% |
|---|
North Lanarkshire | 29.1% | 70.9% |
|---|
East Lothian | 28.4% | 71.6% |
|---|
South Lanarkshire | 28.0% | 72.0% |
|---|
South Ayrshire | 27.3% | 72.7% |
|---|
Dundee City | 27.1% | 72.9% |
|---|
Inverclyde | 25.0% | 75.0% |
|---|
East Dunbartonshire | 24.8% | 75.2% |
|---|
Clackmannanshire | 24.4% | 75.6% |
|---|
West Lothian | 23.1% | 76.9% |
|---|
East Ayrshire | 20.0% | 80.0% |
|---|
West Dunbartonshire | 19.4% | 80.6% |
|---|
Fife | 17.8% | 82.2% |
|---|
Angus | 17.2% | 82.8% |
|---|
Aberdeenshire | 16.9% | 83.1% |
|---|
Moray | 16.7% | 83.3% |
|---|
Perth & Kinross | 16.7% | 83.3% |
|---|
Stirling | 16.4% | 83.6% |
|---|
Highland | 15.4% | 84.6% |
|---|
North Ayrshire | 15.3% | 84.7% |
|---|
Scottish Borders | 12.5% | 87.5% |
|---|
Eilean Siar | 10.7% | 89.3% |
|---|
Dumfries & Galloway | 9.4% | 90.6% |
|---|
Argyll & Bute | 8.3% | 91.7% |
|---|
Shetland Islands | 2.9% | 97.1% |
|---|
Orkney Islands | 0.0% | 100.0% |
|---|
Total | 25.4% | 74.6% |
|---|
Notes to table
Source: Scottish Household Survey 2003-4 (Authors' analysis)
Modelled/synthesised delay
2.4 The congestion mapping utility in the Transport Model for Scotland (Lumsden, 2005) compares freeflow travel times (as defined in the core network coding of the model) with capacity restrained travel times (when the model is in equilibrium) for each link and turning movement in the network. The resultant delay per veh-km for each section of the network (500m grids) is calculated and plotted through a GIS system. As can be seen from Table 3.2 the urban local authorities of Glasgow, Edinburgh and Aberdeen have the largest proportion of their road network subject to more than 0.3 mins 1 of delay per veh-km. Road links with more than 0.3 mins delay per veh-km form the top 10% of links with the most delay per veh-km. As can be seen from Figure 3.1 to Figure 3.5 these delays are not spread uniformly across the road networks of these authorities. In Glasgow the areas experiencing the most delay per veh-km are in the city centre near and along the M8, along the arterial routes through the West End and in pockets on the south side. In contrast, almost all the largest delays per veh-km in Aberdeen City occur in the city centre, whilst in the City of Edinburgh the largest delays are scattered throughout the city including the city centre, the western arterials (A8 and A90), in the suburbs on both the south side (Gorgie Rd, Slateford Rd and Morningside) and the north side ( e.g. Ferry Rd, Granton and Newhaven) as well as on the Forth Road Bridge.
2.5 A review of the 'congestion mapping' figures (Kocak, 2005) identifies that delay 'hotspots' occur principally in urban areas and that the trunk road network contains very few hotspots - as identified using this delay indicator. Sections of the trunk road network which this analysis indicates have 'low' levels of congestion, but where congestion is typically viewed as a problem include: the A8 and A80 in North Lanarkshire (see Figure 3.5), the A720 (Edinburgh City Bypass) (see Figure 3.2) and the M90 (just north of the Forth Road Bridge). This peculiarity is attributed to the nature of the delay indicator, as substantial reductions in speed on the motorway network (from 70 mph down to 45 mph) are required to generate a delay in excess of 0.3 minute per veh-km, whereas much smaller reductions in speed are required to generate the same delay on the urban road network (from 30 mph down to 24 mph). Clearly this raises an issue regarding the most appropriate indicator(s) with which to measure congestion - this is discussed more fully in the following chapter.
Table 2.2 - Proportion of road network subject to congestion by Local Authority ( AM peak trips only)
Local Authority (2)(3) | Proportion of road network (1) in the 2002 AM Peak by local authority subject to an average travel time greater than free-flow speed by: |
|---|
> 0.4 mins/veh-km | > 0.3 and
<= 0.4
mins/veh-km | > 0.2 and
<= 0.3 mins/veh-km | > 0.1 and
<= 0.2 mins/veh-km | > 0 and
<= 0.1 mins/veh-km | 0
mins/veh-km |
|---|
City of Glasgow | 49% | 6% | 6% | 10% | 25% | 4% |
|---|
City of Edinburgh | 40% | 5% | 6% | 11% | 23% | 14% |
|---|
City of Aberdeen | 33% | 4% | 3% | 15% | 43% | 2% |
|---|
City of Dundee | 5% | 10% | 6% | 23% | 33% | 23% |
|---|
North Lanarkshire | 7% | 5% | 4% | 17% | 55% | 13% |
|---|
East Dunbartonshire | 7% | 2% | 5% | 14% | 53% | 19% |
|---|
Falkirk | 5% | 1% | 3% | 17% | 63% | 11% |
|---|
East Renfrewshire | 6% | 0% | 5% | 6% | 70% | 12% |
|---|
Midlothian | 4% | 2% | 1% | 8% | 71% | 14% |
|---|
South Lanarkshire | 5% | 1% | 4% | 9% | 72% | 9% |
|---|
Renfrewshire | 4% | 1% | 4% | 9% | 70% | 10% |
|---|
Fife | 3% | 1% | 2% | 8% | 77% | 8% |
|---|
Stirling | 2% | 1% | 3% | 8% | 80% | 6% |
|---|
North Ayrshire | 2% | 1% | 1% | 3% | 82% | 12% |
|---|
Inverclyde | 1% | 1% | 0% | 2% | 81% | 15% |
|---|
Clackmannanshire | 1% | 1% | 2% | 15% | 69% | 12% |
|---|
East Lothian | 2% | 0% | 2% | 2% | 70% | 25% |
|---|
West Lothian | 1% | 0% | 2% | 5% | 83% | 8% |
|---|
East Ayrshire | 1% | 0% | 1% | 3% | 91% | 4% |
|---|
Perthshire & Kinross | 1% | 0% | 0% | 1% | 85% | 13% |
|---|
South Ayrshire | 1% | 0% | 2% | 3% | 87% | 8% |
|---|
Dumfries & Galloway | 0% | 1% | 0% | 1% | 77% | 21% |
|---|
Aberdeenshire | 0% | 0% | 1% | 4% | 75% | 21% |
|---|
The Borders | 0% | 0% | 0% | 0% | 90% | 10% |
|---|
Angus | 0% | 0% | 0% | 3% | 84% | 13% |
|---|
West Dunbartonshire | 0% | 0% | 4% | 6% | 66% | 24% |
|---|
Total | 8% | 2% | 2% | 7% | 69% | 12% |
|---|
Notes
1. Proportion of road network represented in the Transport Model for Scotland (principally roads that have more than just a local function).
2. Excludes the 6 local authorities not full represented in the TMfS: Argyll and Bute, Highland, Moray, Shetland Islands, Eilean Siar and Orkney.
3. Rank based on the proportion of links with a travel time greater than 0.3 mins/veh-km over free-flow travel times
Notes to table
Data source: Robinson (2006). Data analysis: authors
Figure 2.1 - Glasgow City Congestion map (2002 AM Peak)
Source: Kocak (2005)
Figure 2.2 - City of Edinburgh Congestion map (2002 AM Peak)
Source: Kocak (2005)
Figure 2.3 - Aberdeen City Congestion map (2002 AM Peak)
Source: Kocak (2005)
Figure 2.4 - Dundee City Congestion map AM Peak
Source: Kocak (2005)
Figure 2.5 - North Lanarkshire Congestion map AM Peak
Source: Kocak (2005)
Congestion on Scottish Trunk Roads 2003
2.6 The report Congestion on Scottish Trunk Roads 2003 (Scottish Executive, 2005) uses traffic count data to develop congestion indicators for 22 routes on the trunk road network throughout Scotland. These locations have been chosen to include "those sections of the network which currently experience congestion or which are thought likely to [experience congestion] over the coming years". The locations are set out in Table 2.3 and Table 2.4. The first thing to note is that with the exception of four routes these locations are all in urban or peri-urban areas. Of the four exceptions three are estuarial crossings and the other is the M8 between Glasgow and Edinburgh. From this it can be seen that the sections of the trunk road network that policymakers perceive to have, or will have, a congestion problem are those where traffic flows are heavily influenced by the urban environment or where capacity bottlenecks potentially exist (estuarial crossings).
2.7 The previous analysis identified that trips with destinations in Glasgow, Aberdeen and Edinburgh are those that are subject to the most delay. Table 2.3 and Table 2.4 provide further confirmation of this position for Glasgow and Edinburgh, in that it is sections of the trunk road network that are closest to these cities that have the largest proportion of traffic experiencing speed reductions and journey time reliability problems. The localised nature of these speed and reliability impacts can again be observed as it is only certain routes in an area and certain directions of travel that experience the worst of these impacts. For example the A90 through Dundee in an eastbound direction has 10% of the vehicles experiencing a speed reduction of more than 25% of the freeflow speed, whilst 0% of vehicles experience such a reduction in the westbound direction. A similar effect can be seen on the approaches to the Kincardine Bridge.
Table 2.3 - Speed reduction on trunk roads by route
Area | Trunk road route | Speed: %age of vehs with speed reduced by more than 25% of freeflow speed |
|---|
Direction-1 | Direction-2 | Average for both directions |
|---|
Glasgow | M8 St James Int to Baillieston Int | 11% | 13% | 12% |
|---|
Glasgow | M80 Steppes Bypass / A80 to M80 J4 | 7% | 7% | 7% |
|---|
Kincardine Bridge Approaches | A977 (Gartarry Rbt)/A985 (Longannet)/A876/M876 to M9 Jcn 7 | 12% | 2% | 7% |
|---|
Edinburgh | A720 City Bypass from A1 to M8 | 7% | 7% | 7% |
|---|
Forth Bridge Approaches | A92 Cowdenbeath Jcn amd M90 Jcn 4 to Forth Road Br | 7% | 4% | 5% |
|---|
Dundee | A90 Inchture to A90 Forfar Rd | 10% | 0% | 5% |
|---|
Glasgow | A725 | 5% | 5% | 5% |
|---|
Glasgow-Edinburgh | A8/M8 Baillieston to Hermiston Gait | 6% | 3% | 5% |
|---|
Edinburgh | M9 from M8 Claylands to M9 Spur | 3% | 4% | 3% |
|---|
Aberdeen | A96 Muggiemoss Rbt to Blackburn | 3% | 3% | 3% |
|---|
Glasgow | M77 Greenlaw Jcn to M8 Jcn | 5% | 1% | 3% |
|---|
Aberdeen | A90(N) Balmeddie to Muggiemoss Rbt | 1% | 5% | 3% |
|---|
Ayrshire | A77 Dalrymple to Dutch House Rbt | 1% | 4% | 3% |
|---|
Glasgow | M73 to M74 J7 | 2% | 2% | 2% |
|---|
Ayrshire | A78 Stevenson to Dutch House Rbt | 2% | 2% | 2% |
|---|
Aberdeen | A90(S) Muggiemoss Rbt to Stonehaven | 1% | 2% | 1% |
|---|
Edinburgh | A1 Macmerry to A720 Jcn | 2% | 0% | 1% |
|---|
Dundee | A90 Forfer Rd (Tealing) via Tay Br to Forgan Rbt | 1% | 0% | 1% |
|---|
Erskine Bridge | A898/A898 | 1% | 1% | 1% |
|---|
Perth | M90 Bridge of Earn to Broxden and Friarton | 0% | 1% | 0% |
|---|
Perth | A9 from junction B934 to Luncarty | 1% | 0% | 0% |
|---|
Ayrshire | A77 nr Fenwick to Dutch House Rbt | 0% | 0% | 0% |
|---|
Notes to table
Source: Scottish Executive (2005); Authors' analysis.
Table 2.4 - Reliability (journey time variability) on trunk roads by route
Area | Trunk road route | Reliability: %age of vehs with journey time 15% longer than average for that period. |
|---|
Direction-1 | Direction-2 | Average for both directions |
|---|
Glasgow | M8 St James Int to Baillieston Int | 9% | 11% | 10% |
|---|
Kincardine Bridge Approaches | A977 (Gartarry Rbt)/A985 (Longannet)/A876/M876 to M9 Jcn 7 | 9% | 3% | 6% |
|---|
Glasgow | M80 Steppes Bypass / A80 to M80 J4 | 6% | 5% | 6% |
|---|
Edinburgh | A720 City Bypass from A1 to M8 | 5% | 5% | 5% |
|---|
Glasgow | A725 | 6% | 4% | 5% |
|---|
Aberdeen | A90(N) Balmeddie to Muggiemoss Rbt | 2% | 6% | 4% |
|---|
Glasgow-Edinburgh | A8/M8 Baillieston to Hermiston Gait | 4% | 3% | 3% |
|---|
Edinburgh | M9 from M8 Claylands to M9 Spur | 4% | 3% | 3% |
|---|
Forth Bridge Approaches | A92 Cowdenbeath Jcn amd M90 Jcn 4 to Forth Road Br | 4% | 3% | 3% |
|---|
Glasgow | M77 Greenlaw Jcn to M8 Jcn | 4% | 2% | 3% |
|---|
Aberdeen | A96 Muggiemoss Rbt to Blackburn | 3% | 2% | 2% |
|---|
Glasgow | M73 to M74 J7 | 3% | 2% | 2% |
|---|
Aberdeen | A90(S) Muggiemoss Rbt to Stonehaven | 1% | 4% | 2% |
|---|
Erskine Bridge | A898/A898 | 1% | 2% | 2% |
|---|
Ayrshire | A77 Dalrymple to Dutch House Rbt | 1% | 3% | 2% |
|---|
Dundee | A90 Inchture to A90 Forfar Rd | 2% | 1% | 1% |
|---|
Ayrshire | A78 Stevenson to Dutch House Rbt | 1% | 1% | 1% |
|---|
Dundee | A90 Forfer Rd (Tealing) via Tay Br to Forgan Rbt | 1% | 1% | 1% |
|---|
Perth | M90 Bridge of Earn to Broxden and Friarton | 0% | 1% | 1% |
|---|
Perth | A9 from junction B934 to Luncarty | 1% | 1% | 1% |
|---|
Edinburgh | A1 Macmerry to A720 Jcn | 1% | 1% | 1% |
|---|
Ayrshire | A77 nr Fenwick to Dutch House Rbt | 1% | 0% | 1% |
|---|
Notes to table
Source: Scottish Executive (2005); Authors' analysis.
2.8 To summarise, there is only a limited availability of literature on the locations of congestion in Scotland. Notwithstanding that, a number of data sources exist that contain information on the impacts of congestion (delay, speed reductions and reliability problems). The information that does exist does not define congestion per se, nor does it define the point at which congestion is perceived to be a problem. In describing the locations where congestion exists in Scotland the approach has therefore been to describe the locations where the impacts of congestion are greatest. No commentary is given as to whether this level of impact is perceived to be a problem, as the data analysed does not contain such information.
2.9 From the analysis of the available data a broad picture emerges. Whilst at the national level only a minority of trips (11.5%) are affected by congestion, this figure disguises large geographic, temporal and journey purpose variations. Congestion impacts are largest in the cities of Glasgow, Aberdeen and Edinburgh (where up to 42% of AM peak travellers experience congestion related delay and up to 49% of the AM peak network generates delays). The trunk road network that experiences the most congestion is that in the vicinity of these cities as well as on the approaches to the Forth estuarial crossings. The peak hours are more congested than the off-peak and commuting and business related trips are more affected by congestion than trips for 'other' trip purposes (there is no data on the impact of congestion on freight movements). Congestion is not however just confined to Aberdeen, Glasgow and Edinburgh and their vicinity as congestion related delays are reported throughout Scotland, it is just that their frequency and incidence is higher in the large cities - ultimately it only takes one over-capacity junction to impose a congestion related delay on travellers.