Jeremy F. Parker
rev 1, 12 July 1998
I realize that, for the moment, both broad policy and specific implementations have essentially been fixed. However, I raise my concerns anyway, where the Design Manual's text brings out the issues. The concerns that I, and other cyclists, have will not go away, even though the Design Manual has been declared final. Indeed, after the Design Manual begins to circulate publicly, and more paint begins to appear on the road, voices will become more vociferous. It is better to face the issues as soon as possible.
I believe that the decision to issue such a document, to emerge when the London Cycle Network was already one third complete, shows that the bureaucrats know that there are severe problems with what they have built so far.
I believe that the London Cycle Network, if built to this plan, will decrease rather than increase cycling. A cycle network may encourage large numbers of cyclists to ride their bikes once. That ride will not encourage them to ride twice.
I note the delays of the bureaucrats in organising any kind of monitoring of the effects of building the LCN facilities. I believe that they suspect that the results of any such study would not be flattering. However, I believe that such monitoring is vital to show what actually does work, and what does not.
I believe that, even with a good Design Manual, the bureaucrats of London are not to be trusted to be capable of building good bike facilities. And this is not a good Design Manual.
Furthermore I believe that idea of a London Cycle Network will inevitably harm cycling and cyclists
Furthermore I believe that the cost benefit study for the London Cycle Network is a ludicrously implausible document. Using the framework of that document as it stands, but modifying the numbers it uses to the truth, would, I think, result in a statement that the network does not have a positive cost benefit. However, the cost benefit study does not consider all factors. I am curious as to what a good cost benefit document would show. It might give a result either way
Detailed comments on the Design Manual follow.
The last rationale for route planning was the old GLC's which was based on allocating the most money where there were the most cyclists. In other words, find out what cyclists like, and then make it different.
The phrase "a robust approach to providing..." is a curious one to use. It leaves a rather Stalinist impression, or perhaps just an impression of macho posturing.
1.1.6 sounds like a proposal to sell the London Cycle Network to other road users as a means of getting cyclists off the roads. Such a policy is highly dangerous to cyclists, because, when cyclists fail to use poor facilities, there are likely to be calls for use of such facilities to be made compulsory, and likely to be attempts by motorists to harass the cyclists onto the facilities.
1.2.2. "Access" is this year's buzzword in planning, but is ambiguous, and causes confusion.
Access is currently most often used in the context of the changing geography of Britain, because of such things as out of town shopping centres. This is not usually relevant to London. All of London is already built up, or else unbuildable on, being the Green Belt or similar. The few exceptions are major, high profile, projects, such as the Kings Cross Goods Yard. In such projects many parties are involved and individual project plans will exist. Because such projects are essentially "greenfield" the "some compromises must be made" philosophy of this Design Manual is not appropriate there.
Two access problems do exist in London, and one does not. There is no "last mile" access problem, no problem in getting up the driveway. However, there is a universal problem of finding safe and convenient parking once one has reached the destination. Furthermore cycling has to deal with the sheer size of London, and with competition from other travel modes.
Throughout the world mass cycling only takes place when trips are less than about 3 km long. Such distances form a very small portion of London's required trips. It is essential to encourage cyclists to take longer trips. Such encouragement has even more benefits than at first apparent, because the area reachable by bicycle, and thus the number of destinations reachable by bicycle, goes up not just proportionally to the acceptable trip length, but as its square.
To encourage cyclists to ride longer distances routes must be direct, routes must be fast, and routes must be flat. The old turnpike routes are ideal. The same constraints operated on stagecoaches as operate now on cyclists. Furthermore, to overcome London's size problem, cyclists must be able to achieve as great a distance as possible for a given effort. Cyclists must be encouraged to keep their chains oiled, their tyres pumped up, and miscellaneous pieces from rubbing. Cyclists must be encouraged to use bikes and to wear clothes which do not waste energy in rolling resistance or air resistance. Cyclists must be given intersections which do not give them long red phases at traffic lights, and which do not require multiple light phases to get through. Cyclists must be given routes between intersections which do not require slowing down to watch for broken glass, car doors, or give way signs, or to constantly be wary of pedestrians unexpectedly stepping sideways.
1.2.4 Cycling is indeed an excellent recreational activity. It makes cycling the only recreational form of commuting. For London we need to encourage the revival of cycling as an urban recreational activity. We need to remind the world that bicycle recreation does not require going to some remote moorland with a mountain bike, but is a pleasant way of wandering round London, seeing the sights. After all, if, for you, cycling is not a pleasant way of wandering round London, then you are doing something wrong, perhaps something that could kill you, and you should get off the road until you discover what mistakes you are making, and fix them. If, for you, cycling is not a pleasant way of wandering round London, you may be teaching your children to kill themselves.
1.2.5 says, "The success of the Cycling Strategy will be therefore dependent, as will other elements of a sustainable transport system, on approaches towards company car and off road taxation, allocation of road space to private cars, road pricing etc." The relevance of this to a manual on designing the London Cycle Network is unclear, but it appears to be a warning that the network, by itself, will fail to encourage cycling.
The discussion on safety is obscure and fails to say how the London Cycle network will help. It points out, for example, the major problem of cyclists being killed because they get on the inside of left turning HGVs, yet a major component of the Design Manual is encouragement of lanes, which seem very likely to encourage cyclists to place themselves in the death zone.
The Design Manual says, "It is also argued that based on analyses of off road facilities in Milton Keynes, that segregated facilities do not reduce chances of being hurt as much as is often thought." Actually, Milton Keynes shows that their facilities seem to increase the chances of being hurt, not reduce it at all. Data from many places shows that sidepaths and riding on the pavement generally seem to triple the hazard, although two way paths are more dangerous still.
1.2.6 describes London's target for increasing cycling. This must be used to set targets for the individual boroughs, and to set targets for cycling increase from construction of individual facilities.
1.2.7 The bicycle trips number looks wrong. Hasn't there been a confusion between `London' and the LATS area?
1.2.8 It's a pity that census data was omitted from these cycling statistics. It surely would have been of interest to note that that between the last two censuses six London boroughs were the areas with the fastest growth of cyclecommuting in Britain. (Cycling GB 96, section 4.5)
This section points out the harm that the proposed "ghetto in the gutter"/lanes everywhere policy will cause, and references Toy's prize winning essay, written for the 1995 Velo City conference, about the bankruptcy of that policy. Unfortunately the very next section, 1.5.2, compels use of facilities that the previous paragraph has damned.
1.5.2 Ah yes, the infamous CROW fig 4.3 (CROW), the militant motorists' friend. CROW was bad enough, and Sustrans (Sustrans, fig 3.1) was worse, but where on earth did the numbers in diagram 1.1 come from. The policy fixed by diagram 1.1 undoes all the good words in section 1.5.1, and instead turns cyclists into the niggers of the road, herding them off into bicycle bantustans and ghettos-in-the-gutter in as many places as the planners could get away with.
Why the "of course" in note 2. When a university lecturer uses that phrase, it's always the sign of a weak point in his argument that he knows he cannot prove.
This graph implies that every country lane round London, even the ones listed in cycling guide books, even the ones used on LCC rides, are actually unsuitable for cycling, and so, something must be done to change them
It is possible that this graph confuses traffic calming with traffic restraint. If the graph actually means what it says, and traffic calming is introduced under the conditions shown, there will be a lot of defeated councillors in the subsequent elections.
This graph quantifies an anti bicycle plan that has never, anywhere in the world, been shown to work, and which, in London is already proving unpopular. .
There should be a cross reference here to section 8.7.1, which describes the debris problems of lanes
1.5.3 This section is an example of the "magic marker madness" so prone to affect planners. It has three problems. The first problem is the assumption that universally something must be done. This is likely untrue in most places, and yet nowhere is the "do nothing" alternative suggested. Many people cycle in London, and between the last two censuses inner London was the fast growing area of Britain for cycling (Cycling GB 96 section 4.5). The roads as they are are attracting numerous cyclists, and the no build solution is always cheapest and quickest.
The second problem is that the Design Manual assumes the planners' task is not to make spot improvements where cycling conditions are bad. Frittering money away along the whole of a "route" is likely to waste most of it where improvements are not needed, and thereby leaving only enough to produce inferior, or no solutions to the problems that really exist.
Thirdly the Manual assumes that the task is to produce "a network", thereby defining nine tenths of London's roads as not part of the network for cyclists. Route choice is always very personal, and with everybody's choice being different every cyclist but one will look at the planner's choice and decide that the planner is an idiot (It is often said that cycle planning is always "by idiots, for idiots") Furthermore, when all the cyclists look at their chosen route, and see that London has defined it as not being a cycle route, they will conclude that the london cycle Plan is a plot by motoring dominated planning departments to neglect their obligation to maintain all streets in a suitable state for cycling.
1.5.5 Examples are actually on previous pages
1.6.1: Table 1.6.1 fails to allow for the fact that cyclists need to lean over on curves. Because of the extremely tight curves so often used in Britain cyclists will need to lean over even more here than they would on facilities in other countries. See my comments to section 1.6.6
The tables of 1.6 fail to allow properly for tricycles and trailers. what seems to have happened is that data has uncritically been taken from Dutch sources, where all except narrow cycles are banned. In diagram 1.6 at least 1.2 metres, not 1.0, is needed for a cyclist
. 1.6.2 Who is Babtie? Why is he not referenced? Where did the listed speeds come from? What is the point of listing lower maximum speeds at all, when section 1.6.1 has already pointed out that all facilities must accommodate the faster cyclists?
1.6.3 The width figures are inaccurate in table 1.6.3, which seems to have been taken unchanged from Dutch sources, without allowing for the fact that wider designs of cycle are illegal in the Netherlands, and so the maximum width that will operate there is not necessarily the widest that will operate in the UK
1.6.6 The statement that there is no data on how cyclists make turns is ludicrous. How cycles turn has been known for over a hundred years. See, for example, Sharp, "Bicycles and Tricycles", 1896, reprinted by MIT, 1977. (Sharp) See chapter XVII. The simplest way to incorporate the required turning data is simply to adopt the US design standards, and the reasoning behind them, found in AASHTO's Guide for the Development of Bicycle Facilities (AASHTO)
2.1.3 If this paragraph means that when bikes are less than 5% of the traffic, then bike lanes should never be considered, then I heartily concur. If bikes are more than 5% of the traffic then they will probably be overtaking each other fairly often, and so not confining themselves to the bike lane.
2.1.5 To the sentence "However, any additional lanes may be 3.0m wide minimum as they would not normally be expected to carry cyclists" should be changed to ".....normally be expected to carry cyclists except for those cyclists preparing to turn right, or overtaking obstructions"
2.1.8 The statement in this section that only some cyclists will wish to position themselves in the correct lane when approaching a traffic light is, alas, all too true. This points out the need for education to minimise the number of cyclists choosing the incorrect lanes, and for road markings that discourage all cyclists from blindly sticking to the curb.
2.1.13 I wish that the Borough of Barnet would follow this paragraph. I find their raised rib markings on an advisory bike lane line on the A1000 highly objectionable
There should be a cross reference to section 8.7.1, which describes the debris problems of lanes
Table 2.2 fails to show to what the one-asterisk footnote about substandard facilities applies (most of the table, it seems) The idea that a substandard cycle lane is acceptable where a road narrowing exists is particularly repugnant. That is the last place where cars should be encouraged to force cyclists into the curb.
2.2.1 I find the theory that a bike lane helps motorists to see cyclists better highly implausible, and note that only the opposite effect has ever been detected by experiment (Florida/Wilkinson) the Design Manual should give a reference to the research, if they know of some evidence that what they claim exists really does exist.
The statement that lanes encourage motorists to drive further from the curb is true only when cyclists are not present. In such a case the induced motorist behaviour is harmful to cyclists, because it prevents the sweeping action of the motorists' car tyres from keeping the lane clear of debris. Instead, the sweeping effect only sweeps debris into the lane, not out of it. When cyclists are present the lane stripe induces exactly the reverse effect from that claimed, and this too is harmful to cyclists. It has been shown that, when cyclists are present, motorists drive closer to the curb, not further away, because motorists position themselves according to the lane line, not according to the cyclist. Motorists drive about 15 cm closer to the curb when a lane line is present, and so 15 cm closer to the cyclist. (Florida)
The statement Kerbside lanes, ..., may encourage motorists to drive further from the kerb..." should be replaced by, "Kerbside lanes,whether advisory or mandatory, encourage motorists to pass cyclists about15 cm more closely than would otherwise be the case. However,they are popular with both motorists and inexperienced cyclists, because both groups believe that, with a lane line, they can ignore the traffic in the adjacent lane. Motorists do avoid the cycle lane even when no cyclists are present, thus improving conditions for pedestrians, but preventing the sweeping effect of the cars tyres, see section 8.7.1 of this Manual."
The statement about parking is obscure. Perhaps it is trying to say that there is less parking, legal or illegal, on top of a bike lane than would otherwise be the case. The sentence should be rewritten to clarify its meaning, and if there is any evidence of what they claim, a reference should be given. if the bike lane designer believes that legal parking on top of the bike lane is a bad thing, then no such parking should be permitted.
It should also be pointed out that if parking, legal or illegal, on top of a bike lane does take place, then the lane is harmful, not helpful, to cyclists, because the lane destroys the cyclists' right of way to pull out to pass the parked car. It should also be pointed out that this harm still occurs if the bike lane is temporarily interrupted at the parking place to legalise the parking. [see, for example, St John's St, Islington, for an example of how not to build a bike lane].
The last sentence in section 2.2.1 should be replaced by "However, introducing lanes alongside parked or loading vehicles will tend to result in cyclists getting "doored" unless great care is taken."
2.2.2 A statement should be added here to forbid the all too common practice of fake mandatory lanes. Too many anti cycling planners seem to add solid edge lines and coloured paving to advisory lanes, thereby encouraging motorists to behave as if the lane is a ghetto-in-the-gutter which cyclists should be harassed into not leaving.
2.2.4 A mandatory lane that is not a lane, as would occur in the designed proposed in this paragraph, makes the hazards of the "channelization problem" especially severe. The statement that lanes "cease to be of assistance to cyclists" is a highly rose coloured view of the true harm that part time lanes generate.
A lane which becomes a parking lane at off peak times will create difficulties in cleaning it, and make the well known known problem of bike lanes collecting litter even worse.
Placing a bike lane in between loading bay and curb will result in use of the bike lane for temporary storage of goods being loaded and unloaded, thereby blocking cyclists. It will also result in blocking of the lane outside the loading bay by cyclists (like me) who will be riding there instead of in the bicycle bantustan
2.2.5 The Design Manual says that a 1.5 meter wide bike lane can look like a parking bay. This, no doubt, is because it is painted green, has large bicycle stencils on it, and is about the same length as a parked car [again see St John's St, Islington for an example]. This statement also tells us that the planners do not expect proper parking enforcement on bike lanes any time soon, which implies that lanes should probably not be built in such places before it is known that a reform, and a permanent reform, will be implemented.
The last paragraph of section 2.2.5 talks about competing uses. Why not have an arrangement where the limited resource can be shared among the uses, differently from moment to moment.
2.2.6 Trying to divide 4m of space into two lanes is already bringing the cycling officers of London into discredit and contempt. The idea should be stopped at once.
2.2.7 The idea that a pinch point can simultaneously be narrow enough to pinch, and wide enough for two vehicles to pass side by side is essentially illogical. I always take great care to avoid the bike lane in such places, and to block the general lane to any vehicles behind me. Sometimes I get hooted at Then I slow down and start coasting.
2.2.8 What an inelegant piece of planner-speak, trying to say, "Put the bike lane on the uphill side". When there are hills, bikes going downhill may well be going as fast, or faster, than the other traffic, and so will be avoiding any bike lanes that exist.
2.2.9/2.2.10/table 2.2.9. My comments to para 2.2.6 about the stupidity of trying to divide 4m into two lanes apply even more to dividing 3.5 m of space into two lanes.
2.2.11 The lack of comment about correcting motorists' misunderstanding of the meaning of "mandatory" lanes makes it seem that the Manual is trying to imply that planners should install such lanes in order to take advantage of such misunderstandings.
This paragraph should point out the disadvantage of mandatory lanes: cars don't go in them. This means that the lanes never get swept of debris.
Actually, I do not understand what is wrong with letting cars into a bike lane. After all, they will only do this when no bikes are around. Motorists generally avoid objects close to two metres tall and weighing around eighty kilos, if they are right where the motorist normally looks.
2.2.12 " maintaining a degree of continuity" of a lane across a junction will result in yet another instance of the "channelization problem" - cars and bicycles doing the wrong thing at intersections. Cars will turn left from one lane out from the curb, not from close to the curb, and will tend to cut across cyclists, causing the well known "left hook" problem. This seems a way to cause lanes to induce some of the intersection problems that make bike sidepaths more dangerous than riding in the roadway.
Artic's, of course, have to turn left from a fairly far out position. Putting in a bike lane will encourage cyclists to commit suicide by getting alongside turning lorries, right into the "coffin corner" - already a problem that causes a nearly about a fifth of the cyclist deaths in London.
2.2.13 Taxi drivers are very much better than other drivers in London. Any facility which requires even the best drivers to have a special "code of good conduct" to make it function probably has some inherent design flaw, as indeed the lanes proposal does.
2.2.14. It does not matter if a lorry occasionally drifts into a bike lane. That will help to keep the lane swept clear of debris.
2.2.15 An advisory bike lane is advisory usually because it is of substandard width. The "channelization problem" therefore becomes more serious, and it is especially important not to emphasize such lanes by colour.
2.2.16 This paragraph seems designed to encourage cars to turn in a manner which is, quite rightly, illegal in the USA (USA), although Britain's Highway Code is merely confused on the subject.
2.2.18 Of course it's possible to get bike lanes past bus stops, but that doesn't mean it's a good idea. If the bike lane goes behind the bus stop, then I will leave the bike lane, and pass in front of the bus stop.
2.2.19 Where bike lanes deflect, I take great care to deflect out of the bike lane, into the next lane, well in advance.
2.2.22 The idea of stopping a bike lane 5 metres in advance of a junction to allow bikes to change lanes is ludicrous. The cyclist should be changing lanes more like 50 metres in advance than 5. Actually the American "Manual on Uniform Traffic Control Devices" (MUTCD) recommends 15 metres, although this is widely regarded as too short. California standards require 30 metres. (BF26)
2.3.1 This section talks about putting contra-flow lanes in gyratory systems. This is essentially equivalent to putting a contra-flow lane round the central island of a roundabout - gyratory systems are essentially big roundabouts, after all. This means that gyratory systems are about the one kind of place where contraflow lanes are probably not a good idea.
2.3.8 Block paving - ugh. Why discriminate against skinny tyres. I don't want to train for the Paris-Roubaix race, the "hell of the north", that such paving will imitate.
2.3.9 It is interesting that `protect' at the bottom of p37 is in inverted commas. We should have more such inverted commas in this Design Manual, and perhaps fewer designs where such inverted commas have to be used.
On independent alignments, the key distinction is between shortcut `back alley' routes, and longer Sustrans type paths
2.6.3. Sidepaths going with traffic are about three times as dangerous as the adjacent roadway. Sidepaths going against traffic are about 12 times as dangerous (ADFC), (Dutch). Because of the general unacceptability of 2-way sidepaths, wording similar to that in section 12.1.2 of cycle-friendly Infrastructure should be added here.
Section 2.6.3 talks about "one-way tracks". In real life there is no such thing. they get used as two-way tracks, and must be designed accordingly. 50 cm clearance to posts is specified. The USA (AASHTO) requires 60 cm. The Dutch (CROW, p84) require 75 cm.
2.6.4 specifies 50 cm minimum between cycle track and carriageway. In the USA (AASHTO) triple this is required.
2.6.5 talks of "waiting areas" for motorists, implying that they must give way to cyclists as a cycle track crosses side streets. This is good, if some way can be found of making sure the motorists actually do give way as legally required. However, elsewhere in the Design Manual it seems that the cycle tracks will rarely, if ever be built to give cyclists priority. Witness the photograph on p53 of the Design Manual, which shows cyclists having to give way even at a driveway
2.6.10 Instead of saying "Such facilities can prove extremely attractive..", why not say, "Good facilities will ..."
2.6.16 Putting bollards on fast downhill stretches of bike trails is a good way to cause accidents (Sustrans?) , see 2.6.18 of the Design Manual.
2.6.17 This should be "Motor vehicles can be..." Bikes are vehicles too. Once again we see the prejudice of the planners in the language they use.
2.6.18 - 2.6.19 The barriers are the most disliked features of Sustrans paths
This section fails to make clear the great difference in problems between putting bicycles on the pavement, and putting bikes on an independent trail
2.7.3 "Such ideas can be contentious". Indeed so, and we should try to increase the contentiousness whenever possible. I will do my best.
2.7.6 Unsegregated facilities are the norm in the USA, as are the associated problems. It is generally assumed that, whatever the original width, 2' (60cm) of extra width will solve everything.
The solution recommended in the last sentence of 2.7.6, giving the pedestrian a wide berth, alas, is not possible. We are told that the sidewalk is less than 3 m wide; there is no room for a wide berth. Bike bells help, but a voice call does not work as a substitute for a bell. Whistles work against cars, indeed so well as to often be banned in the USA, because they remove some of the cyclists' disadvantages. but I have not tried a whistle against a pedestrian. Air horns work even better than bells and whistles, but are unsporting, no matter how much satisfaction they give when you see the results (demo' on request).
2.7.8 This section implies that it is sometimes legitimate to assume that visually impaired users will never use the facility, and to design out usability by such users. This seems very similar to the theory that it is legitimate to design out usability of the road system by cyclists, or to design out the crossabilty of roads or railway tracks by pedestrians or cyclists. Surely none of these are acceptable.
2.7.9 This implies that two way paths are acceptable alongside roads. Indeed, a photograph of such a path is shown. This is generally not true, and British standards say that such facilities should not be used. Language similar to paragraph 12.1.2 of Cycle-friendly Infrastructure, or even stronger, should be added. Note my comment to 2.6.4
The sentence talking about capacity versus width implies that minimum widths are set by expected capacity, and that therefore on lightly travelled routes a substandard design is acceptable. This is not true, and the sentence should be removed.
2.7.10 The statement that "Pedestrians and cyclists can be segregated by [a] solid white line" is false. As anybody who actually rides on such facilities can attest, nobody stays on the "proper" side of their line
2.7.11 This implies, falsely, that less shy distance is needed to a railing than to a curb. In fact the situation is just the reverse. See CROW, 4.3.1.3 for details.
2.7.15 The photo on p53 shows a "rehabilitated" 1930s cycle track, made even more dangerous by permitting cyclists to go the "wrong" way.. I note that cyclists have to give way even at driveways. I 'm not sure that it is wise to show photos like that if we want to encourage cycling in Kingston.
2.7.18 This section required 0.25m because of the railing. US standards (AASHTO) would require 0.60m
3.1.9 The idea of a lane to "help guide the cyclist to the nearside lane" is very bizarre. Why should cyclists need "help" to carry out such a task? It sounds like one more anti cyclist idea to "get the goddam cyclists out of the motorists way." Or would there be some strange way for bikes to turn left?
3.1.10 The idea of "helping" through cyclists by putting them into the "coffin corner" to the left of left turning vehicles seems very dangerous. Given that using this strange facility is liable to cause excessive delay as well as probably some danger, I imagine most cyclists would avoid it, and join the other through vehicles.
3.1.14 The G-turn or jug-handle layouts to "ensure" that I don't make a vehicle right turn will fail. It will take more than that to stop me from avoiding such a design's excessive delays.
3.1.18 Advanced Stop Lines do help right turning cyclists position themselves correctly at the traffic light, but often counteract that by encouraging them to position themselves wrongly as they approach the light. See 3.1.23 - 3.1.25 for countermeasures.
. Note the implicit assumption built into this section that cyclists must always give way to motor vehicles, no matter how many cyclists there might be, and how few motorists. (I know of one such intersection, fortunately in Massachusetts, not Britain, where the cyclists have to stop and give way even though the roadway is chained off, so the motor vehicle traffic is normally zero.)
3.3.8 Changing priority to favour bikes travelling along a road is also liable to encourage motorists to use the road.
3.3.9 It is not just convenience that makes me avoid sidepaths, it is safety as well. The fact that the authors of this manual do not even recognise the safety problems, makes it difficult to believe that any facility built under their advice will ever solve those safety problems.
3.3.10 I am sceptical that a facility with different right of way rules for pedestrians and cyclist will ever work. Another reason for me to steer well clear of such things.
3.3.15 The problem of crossing slip roads is made much worse in Britain by the lunatic design of British slip roads. The design would still be crazy even if the bicycle had never been invented, and is just as lunatic in places where no cyclists venture, as on motorways. Slip roads should have a lane parallel to the main lanes to allow entering traffic to accelerate to match the prevailing road speed, and then to change lanes onto the main roadway at a time of their own choosing, when a suitable gap appears. Instead Britain has the triangular, ever narrowing, design which tends to encourage motorists to force their way onto the main road regardless, safe or not. The design of British slip roads is unsafe even when only motorists are involved. Putting such things where cyclists must ride is even worse. Why Britain continues to allow them I cannot imagine.
3.5.3 "Conventional roundabouts with large flared entries and smaller central islands are more dangerous ..." should be changed to "...even more dangerous...". Other roundabouts are still dangerous.
The weak wording of these two sections is an indication of the anti cyclist bias of the whole Design Manual
3.5.9 Grade separation can indeed be better than roundabouts. However, the design standards proposed in section 3.4 are totally inadequate. See my notes to section 3.4.14. American (AASHTO) design standards should be used instead.
3.5.14 Laned roundabouts. This crazy, crazy, idea is justly controversial in Barnet, where they have been introduced. Such designs put the cars in the bikes' blind spots, the bikes in the cars' blind spots, and then make them cross each others' paths. An attempt to do something about this has been to put give way signs all the way round the roundabout on the bike lanes. This only makes the still continuing accidents "the cyclists fault". It also has the peculiar effect that the cyclist can regain his right of way by moving six inches sideways out of the bike lane, because the only location where the cyclist does not have right of way is in the lane provided for him.
If you value your life, stay well away from the bike lanes on roundabouts.
3.5.16 Central island crossings. I have seen the canonical example, at Croxteth Gate in Liverpool. On a quiet Sunday morning, with no cars around, it was usable. I can't imagine myself using it a second time though. I would stick to the roadway with the other vehicles.
Furthermore any attempt to create the idea that signs warn of cyclists is dangerous to cyclists, because it implicitly carries the message that lack of signs tells of lack of cyclists, implicitly giving motorists permission to "not see" cyclists on normal roadways.
Note also the problem of debris, referred to in section 8.7.1
7.2 The idea that signs should be put on lampposts where possible is exactly wrong. Lampposts are the one place where signs become invisible at night. Therefore signs should avoid lamp columns.
7.5.2 research in the USA (Florida) has shown that bike lanes tell motorists that they need not look out for cyclists, because the cyclists are not in the motorists' lane. The result is that the motorists pass the cyclists 15 cm more closely than they would if they remembered to pass with "appropriate consideration".
It would be helpful to give a reference of the supposed slowing influence of advisory bike lanes
7.5.5 Some kinds of thermoplastic paint are slippery. Will these be automatically avoided?
sign [950] Why is this included twice?
7.7.7 Is there any time schedule for boroughs to specify a list of their secondary destinations?
Directional signs. It would be nice to have an example of a bypass sign, e.g. Southwark Br, avoiding Elephant & Castle
More than safety should be audited. At least convenience should be audited, and it would be also good to audit obnoxiousness/pleasantness and anti-cyclist discrimination in audited facilities.
8.7.1 Sections 1.5.2 and 2.2 should have a cross reference to the debris problems of lanes, mentioned in this section.
surfacing: the direction of ridges and slots is critical - transverse ones do not matter so much. Quasi-longitudinal ones are killers, see Forester, "Bicycle Transportation, p251 (Forester 2)
J. Toy should be more fully referenced. Was his essay presented at the 1995 Velo City, in Basel?
A number of items are mentioned in the text, but not referenced. These include:
p 10, LRC "Cycling in London" Aug 1997
p 11, a future DETR report
p 21, "Babtie"
Progress: "A design guide for planning and designing cycle facilities on the network will be published in the summer of 1997." Well, actually, it wasn't. The delay in issuing the Design Manual seems to point to some problems. The fact that even its planned issue date is when about a third of the network is already in place is also somewhat odd. The reluctance to let anybody see the draft seems to show that the bureaucrats are not proud of their document. The perceived need for the document at all seems strange, when other documents exist.
AASHTO: American Association of State Highway and Transportation Officials Guide for the Development of Bicycle Facilities 1991. See the section Separation Between Bicycle Paths and Roadways, and the various curve related sections
ADFC: some information is given in McClintock, The Bicycle and City Traffic, 1992. The internet newsgroup de.rec.fahrrad also carries information, see for example the message by Rasmus Althoff,( title unknown), on 6 July 1997
BF26 Bicycle Forum 26, Nov `90, Reviewing the `81 AASHTO Guide
CFI: Cycle-friendly Infrastructure CTC etc. 1996
CROW: CROW Sign up for the Bike, CROW 1993
CTC 95: D. Matthew, More Bikes - Policy into Best Practice, CTC `95
Cycling GB 96: Gov't Statistical services, Transport Statistics Report, Cycling in Great Britain, HMSO Aug `96
DOT: Ministry of Transport,Transport Advisory Council, Report on Accidents to Cyclists, 1938. Note especially FJ Urry's reservation to paragraphs 14 and 15
Dutch Beukers, Bicycle Facilities of the Netherlands, ITE Conf `77 says "safety problems ... almost insuperable"
Florida: Pein, Wayne, bike Lane/Wide Outside Lane study e-mail, 28 july 1997 on the facilities-n-planning@cyclery.com mailing list, reporting Harkey and Stewart. Evaluation of Shared-Use Facilites for Bicycles and Motor Vehicles, TRR 1578, 1997 Note: the subsequent discussion on the mailing list is vital. see the mailing list archives at www.cyclery.com
Forester 2 John Forester, Bicycle Transportation, 2nd ed, 1994, MIT press
MUTCD Manual on Uniform Traffic Control Devices chapter IX
ratings refs: Elizabeth Winternitz, Using Algorithms to assess the Cycling safety of Selected Arterial Streets October 1981, Pedestrian/pedalcycle safety project of the Urban and Regional Planning Program, University of Hawaii at Manoa, for the Motor Vehicle Safety office, Dept of Transportation State of Hawaii.
and the references in the literature survey within the above paper
Sorton and Walsh, Bicycle Stress Level as a Tool to Evaluate Urban and Suburban Bicycle Compatibility, 1994, Transportation Research Board Record 1438
Epperson, Evaluating Suitability of Roadways for Bicycle Use: Toward a Cycling Level-of-Service Standard, 1994, Transportation Research Board Record 1438.
see also other TRB papers referenced at their web site.
Mark Horowitz, Long post on on-street bikeways, e-mail on 11 Oct 97 on the facilities-n-planning@cyclery.com mailing list, and subsequent discussion. See the archives of the mailing list at www.cyclery.com
Sharp: Sharp Bicycles and Tricycles, 1896, reprinted by MIT 1977
Sustrans:Arup The National Cycle Network, Guidelines and practical details, Issue 1, Sustrans, Feb `96
TAL 7/97: TAL 7/97 Supply and Demand for Cycle Parking, DETR Traffic Advisory Unit
USA: see, for example, the California Motor vehicle Code, Division 11, Chapter 6 "Turning and Stopping and Turn Signals",para 22100- (a), viewable on the internet at www.leginfo.ca.gov/calaw.html
Wilkinson Wilkinson, Clarke, Epperson and Knoblauch "the Effects of Bicycle Accommodations on Bicycle/Motor Vehicle Safety and Traffic Operations" FHWA-RD-92-069. US Federal Highway Administration, 1992