London Underground is now moving record numbers of people. Last December, 4 million people travelled through the underground in one day. This had not been forecast to take place until about 2016. In addition, line upgrades are coming on stream, starting with Jubilee and Victoria lines and then moving onto the Northern line and subsurface. These upgrades will effectively increase the number of trains bringing more people and physically more power into the tunnels.
In developing the programme and ideas for the Cooling the Tube programme the first thing that was undertaken was to try to understand what the current temperatures were on the network.
Looking back through historical London Underground records there was some very detailed information between 1915 and 1916 about temperatures and then again from 1931 to 1936 and 1947 to 1951 when there was a massive focus on tunnel cooling (due to war). After this there is no more data and, although it was known the underground was warm, there was insufficient information available on which to make reasonable assumptions, therefore it was necessary to implement a monitoring programme.
One of the first things undertaken to address this was a comprehensive temperature monitoring regime. Temperature loggers that record the temperature every 15 minutes were placed on 67 stations and there is now approximately 12 months of data available, which has provided a good picture of the trends. One month’s worth of data that is of interest was for three station platforms – Belsize Park northbound, Hampstead northbound and Chalk Farm northbound together with the outside temperature at that time recorded in the summer of 2006. Some of the trends that have been identified include:
- The platform temperatures are remarkably stable. Throughout the course of a day the temperature change is only 1 to 1.5°C
- There is also very little fluctuation in temperature over a month and even on one of the hottest days of the summer (32°C) there was little change in the temperature as the thermal inertia of the system managed to damp the temperature rise down quite well, but during the cooler parts of the month it was still at the same temperature
- There are some differences between platforms. On some of the better ventilated platforms, for example the Victoria line platforms, which have more draft relief the temperature can fluctuate by 5 or 6°C, but typically about 3°C in any given day
The key component of the heat within the system is trains, which is a consequence of the amount of passengers. Depending on the train, up to about 90% of the heat comes from the train and ancillary equipment through braking, drive and mechanical losses with less than 5% contributed by the passengers directly.
With passenger numbers on the Underground continually rising; 40% growth expected by 2030 this will increase the problem as more heat will be entering the system through a greater number of trains and therefore increasing the temperature. The challenge facing London Underground is to keep a pace of this growth and make the Underground manage this growth and make it cooler, more comfortable and thermally safer than it is today. Additional heat is also being added to the system not only from an increase in the number of trains during peak and off peak periods but also the use of air conditioned trains and faster trains to reduce journey times.
So where does that heat go?
The tunnel walls remain the principal heat sink despite decades of heating with almost 79% of the heat going into them although this needs to be capped at this level to prevent further system temperature rises. Approximately 11% is absorbed by piston effects with the remaining 10% due to mechanical ventilation. The 10% ventilation is the targeted area to develop cooling systems to try and reduce the temperature.
But how hot can it get? One of the things that is cited quite frequently is the fact that you can’t transport livestock over 35°c. So if you can’t transport livestock over 35°C why can we transport humans? It’s really not as simple as just one sort of temperature. Firstly, thermal heat strain is really dependent on humidity and much less so on air temperature. But also there is no express standard for humans written down as a number on which you cannot exceed.
There are a number of standards for heat strain but the majority of these are based on an industrial environment and an eight hour exposure time. To assess the effect on passengers is more complicated than this and requires examination of the core temperature within the body to work out if that core temp gets above 38°c, which is when you could start to experience the effects of heat stroke. No model however is able to replicate the environment of travelling on the underground. For example parameters such as acclimatisation, pre-exposure, stress, crowding, train boarding, weight, age, posture, etc. need to be considered. There is also the inability of the body to radiate heat in a crowded train.
To enable an accurate assessment of heat strain some quite interesting correction factors are required for the models to account for crowding. Using these finite difference models a design criteria of 29°C for the system has been determined to protect passengers.
Longer term station ventilation models are also being developed where the webs of cross passenger adits are resolved into a circuit diagram by looking at the hydraulic and thermal characteristics. It is then possible to examine the effect of different train frequencies, different train operating parameters and get interesting temperature data and humidity predictions as a function of time.
The system is going to get warmer so ultimately temperature rise will have to be mitigated and in essence there are two ways of doing this
1) known technologies which we can use at the moment without risk
2) future technologies, which involve train comfort, train energy and station comfort
These will need to be looked at on a line by line basis however they can be divided into four key categories:
1) Operational solutions. This is the most energy efficient, environmentally friendly and cost effective method to stop the heat entering the system. Some ways that this could be undertaken are:
- Examining the timetable. Are more trains necessary during inter peak periods? This requires examination of the business case
- Coasting – allowing train to gradually come to a stop under friction rather than breaking so heavily at stations. This is currently used, but could it be used more? The negative effect of this is increased journey times
2) Train based solutions
3) Tunnel based solutions such as tunnel ventilation, cooling pipes, etc.
4) Station based solutions
The project is about reducing the temperature rise in the underground, however this is dependent on both the train operations and the trains and train systems being used or implemented in the line upgrades.
London Underground is undertaking a systems engineering approach to gradually zero in on an optimised solution. This involves working closely with the supply chain Metronet and TubeLines to help identify options and opportunities to try and reduce temperature rise on line upgrades and identify things that can be changed to try and get a better energy solution.
This is an ongoing process and a number of solutions have already been implemented with many more in development. A few of these will now be discussed.
Options
The Seepage Water Cooling System at Victoria was recently awarded a Carbon Trust Award. There is an existing sump at the District and Circle line at Victoria station, which receives a tremendous amount of seepage water. This is believed to be a combination of the River Tyburn, which has historically flowed through that area, surface seepage, natural flow from river terrace gravels and possibly from leaking water mains. This is a large quantity of water, and is relatively cool all year (up to 16°C during summer), that is being pumped out of the underground system into the sewer every day. A pilot scheme has been developed at Victoria to put another pump into the sump, intercept the water and pump it down to the lower concourse. Cooling units are then installed in the adits with ceiling mounted fan coil units and a temperature reduction of a couple of degrees can be affected. This water is then returned to the sewer. In the future as a part of the upgrade works at Victoria it is planned to develop this system further to increase the cooling capacity at the station.
Installation of new fans is another quick fix that has been developed. However this has been done in conjunction with trying to get the most out of the asset. When the cooling programme commenced in August 2005 the project had 38 fans in operation on the network. There is now over 90 fans running or available for use.
Another method of ventilation being considered is under platform exhaust, which is a particularly effective means of ventilating the railway. A lot of heat that comes into the railway comes from train braking, train undercar equipment therefore if some of this can be captured before it has a chance the influence the environment it is a very effective means of providing cooling. Figure 1 shows a typical example where there are exhaust ports in platform and try and use the under platform void as a duct with an adit connecting two ducts together. Another adit then connects to an existing or new ventilation shaft.
This is a solution that is quite invasive in terms of construction and is not suitable for every station.
Non ventilation methods
Where ventilated solutions are not possible other methods being considered are mechanical cooling or water based cooling systems. One of these is known as Borehole Cooling. Figure 2 shows a schematic of the system. The saturated chalk aquifer provides a large reservoir of water that is about 13°C most of the year. The intention is to lift the water from the aquifer and pump around a cooling coil within a station, using distribution pipes around the station which where possible would be located in under platform voids, platform walls etc, to make them less obtrusive, and then re-inject the water back into the aquifer. At platform level an air handling unit over the platform will be required to deliver cooling to the platform areas.
Re-injection of warmer water back into the aquifer will increase the temperature of the aquifer and may therefore affect the extraction point temperature. Detailed thermo and hydraulic modelling of the aquifer to assess the temperature increase locally has been undertaken. At Stockwell station the modelling has determined that the extraction and injection points should be 200m apart which will be difficult to achieve on many of the underground stations. It is proposed to undertake a full scale trial at Stockwell Station in the future.
Where it is not possible to use environmentally friendly or sympathetic methods such as borehole cooling or seepage water mechanical chillers will need to be used. These are practical and affordable but they do come with an energy penalty and development of units that are more architecturally aware will also need to be considered. Where there are draft relief shafts on the system that are not reaching full functionality some are being retrofitted with fans and exhaust systems. In addition the existing cooling systems are being operated on a seasonal basis. By operating the fans a lot harder the temperature wave in the soil can be pulled forward so that we can store some of the cold in those first few metres of the soil behind the platform and use this to try and moderate our temperature. This is a system that has been used effectively on the Moscow underground system.
Evaporative cooling is another option being considered and a trial is to be undertaken at Gibson Square shaft focussing on:
- Wetted media pad
- Spray washer
- Tunnel spray system
The theory of this method is that as the air evaporates it is taking energy from the air to evaporate and that energy manifests itself as a drop in temperature. By introducing droplets of water into the air in the underground, although there may be a slight increase in humidity, there will be a real decrease in temperature.
Hybrid air conditioning
One of the more interesting train based solutions is a trial of a hybrid air conditioning system. One of problems with train air conditioning systems is that the systems themselves dissipate a lot of heat into to tunnel environment resulting in the tunnels getting warmer and then when the trains stop within tunnels they can cut out very quickly because of no ventilation and temperature rise. On lines such as Piccadilly, where the trains spend a good amount of time outside, a bit of time in tunnels and another good amount of time on the outside of their lines, the theory is to develop a system that generates some cooling on the outside, stores that cooling on the train and then uses the cooling while in the tunnel section. Once outside again the process is repeated. The key to doing this is a technology called a phase change medium which is something in the order of ICE but does not quite freeze at 0°C. The concept is that on the outside a condenser and evaporator is used to pass very cool liquid through the phase change medium, freeze it, then when the train is inside the tunnel, turn off evaporating condenser (that generates the heat) and circulate water through the phase change medium to melt it. A concept design is underway for this and a full scale trial on a Piccadilly line train is planned for 2008.
The Cooling the Tube programme is not a ‘one solution fits all’ programme. It is a large systems engineering project, which looks to reduce heat in the first instance with train operations, then implement a whole raft of infrastructure measures to create taylor-made solutions for each station that are optimised in terms of cost, energy use and deliverability.
Jubilee Line platform Westminster Rising temperatures on London’s underground Under platform exhaust is regarded as an efficient cooling technique and could be used at some of LU’s stations Figure 1 – Section of under platform exhaust Borehole cooling would utilise ground water at 13ËšC from the saturated chalk aquifer beneath London and circulate it around the system Figure 2 – Borehole cooling Diagram of the pad, spray and tunnel spray cooling systems Figure 3 – Pad, spray and tunnel spray cooling systems Hybrid air conditioning is a real possibility for the Piccadilly Line where trains spend a significant amount of time above surface Figure 4 – Hybrid air conditioning Questions & Answers
Graham Walmsley, of Meath County Council, asked whether there was any possibility of using heat exchange systems to use the heat from the underground to heat buildings in London during the winter.
This is something that is being considered by the cooling team. On a couple of occasions it has been close to fruition but there has always been an obstacle in the way. There are many factors in making this work not least co-ordination both in space and in time. On the borehole cooling, passive provision is being made for heat exchangers on the outlet. This would allow a stakeholder to connect into it.
Charith, Civil Engineering Dynamics, had three questions: With regard to the core body temperature information, was NASA research looked at?
One of the models used during assessment was developed by NASA in the early 1970’s.
Was Flywall technology considered where the braking generate flywheels and the gearing then recovers the stored energy?
Regenerative braking is being used and developed further in the system to push power back into the system. The traction power system has been improved and composite conductor rail is being used. LU looked at Flywall in the past but there remain some safety concerns due to the amount of embedded energy that would need to be transported on trains. Newer technology such as capacitors is also currently being considered.
Water temperature gain is there in the water for the borehole cooling. 0.15-0.2ËšC increase. The bigger problem is the temperature increase at the re-injection point, but over a 60 years this is predicted to increase the extraction temperature by 1.5-2ËšC.
John Murphy, Tubelines, was interested in the humidity of the system rather than the temperature. The human body is generally not too sensitive to humidity. Currently the relative humidity is around 40% and it is considered that up to 70% would be necessary before it became uncomfortable. The evaporative cooling options would increase humidity and if used extensively throughout the network could get up to this value easily. It is therefore important that these techniques are used correctly.
Barry New, GCG, was interested in the Environment Agency reaction to re-injection of wastewater back into the aquifer?
The current legislation is that the re-injection temperature should not be more than 10Ëš above the extraction temperature and LU are working within these limits.
Rapporteur: Damian McGirr
