Carbon Monoxide Legal Advice

Deaths from carbon monoxide poisoning generated by domestic heating appliances have received a great deal of publicity in recent years.

The build up of carbon monoxide can arise because the appliance itself is faulty, more often however, it is related to misuse, poor installation or maintenance, blocked and leaky flues or inadequate ventilation. Some of these problems can also arise when the occupant deliberately obstructs the ventilation or flue to prevent draughts.

Regulations already exist concerning the installation and maintenance of most gas appliances in rented accommodation, but no such regulations apply to appliances which utilize other fuels, to portable gas appliances or indeed to owner occupied premises, whatever the fuel.

The Department of Environment, Transport and the Regions, chaired an inter-departmental committee including: the Health and Safety Executive, the Department of Health and the Department of Trade and Industry. This group examined the scope for tightening controls on the safety of non-gas fuelled appliances in rented accommodation, taking into account the specific requirements already in place for gas appliances.

This study was carried out by Metra Martech and is concerned with fatalities caused by poisoning from carbon monoxide from domestic heating appliances in rented accommodation, where the appliance is not covered by the existing Gas Safety (Installation and Use) Regulations 1994.

The primary objectives of the study were as follows:

i)   to determine the number of such deaths annually in each of the years 1987 to 1996.

ii)  to identify the circumstances surrounding each case including:

·         details of the accommodation (private, local authority, institutional . . .)

·         the fuel used

·         the service/maintenance history of the appliance

·         the “overall cause” which led to a build up of carbon monoxide.

iii)  to assess the trend in the numbers of cases and to forecast likely future trends.

iv)  to present the findings in a concise report with the data analysed and cross-tabulated to support the conclusions.

SUMMARY AND CONCLUSIONS

i)   The Gas Safety (Installation and Use) Regulations 1998 (GSIUR) require certain landlords to maintain gas appliances and flues provided for their tenants to use in a safe condition, to have them checked annually for safety by a Council of Registered Gas Installer (CORGI) registered gas installer, to provide a copy of the safety check record to tenants and to keep a record of those checks. Appliances fuelled by non-gas combustible fuels such as coal, solid fuels, wood, oil and paraffin etc are not within the scope of these regulations.

ii)  The question as to whether there is any justification for introducing regulations similar to those for gas appliances - to cover those appliances operating on these other fuels - has been raised. This project was designed to address the question.

iii)  Databases were obtained from the Office of National Statistics (ONS) and the Home Accidents Deaths Database (HADD) of cases involving poisoning by carbon monoxide in the home for the years 1987 to 1996. The databases were examined to select cases where the carbon monoxide may have arisen from non-gas fuels. 288 possible cases were identified for further examination.

iv)  Permission was obtained from almost 100 coroners to inspect their files and further details were obtained on most of the 288 cases. For each case we were seeking to establish:

·         the type of fuel used

·         the rental status of the property

·         the cause of the accident

v)   It was established that 141 of the cases involved non-gas fuels, 134 of which were solid fuels. In addition, there were 17 cases for which the type of fuel was not established.

vi)  Of the cases involving non-gas fuels, it was established that 44 occurred in rented accommodation and 71 in owner occupied properties. The status of properties for the other cases could not be established.

vii) After making adjustments for the cases where fuel type or rental status could not be established, it was estimated that there were 55 fatalities involving non-gas fuelled domestic heating appliances in England and Wales between 1987 and 1996. This is an average of 5-6 cases per year.

viii)There is no significant trend in the number of cases per year, either in rented or nonrented

properties. There is no reason to expect, from the statistics, that there will be any increase in the number of cases. There may be some reduction if solid fuel appliances are replaced by gas fired ones and as safety publicity continues to do its job.

2.

ix) There is a substantially higher accident rate among the older sections of the population with over 50% of cases involving those over 65.

x)  There are substantial regional variations in accident rates which are particularly high in Wales and the Midlands.

xi) There is considerable seasonal variation in accident rates with almost 80% of deaths occurring in the winter months.

xii) The most likely type of non-gas appliance to be involved is a room heater (an enclosed stove burning smokeless fuel) which accounts for 42% of cases in rented accommodation. Central heating boilers account for 20% of cases.

xiii)There are three type of factors which contribute to accidents, these involve the User on a day to day basis, the original Installation or its Maintenance. User factors are the most important. Solid fuels produce fly ash which accumulates in the flue, particularly in the throat (the narrower junction between the appliance and the flue and often incorporating a damper mechanism). It is normally recommended by manufacturers that the throat is cleaned once a month yet almost 40% of fatalities occur, at least in part, because this has not been done. Inadequacies in the original installation, particularly with respect to the provision of permanent ventilation also play a significant part in causing accidents. Lack of maintenance as such is less important.

xiv)       Educating users in the importance of clearing the throat of the flue, having the flue checked if the performance of the appliance suddenly changes and using heaters with the fire door closed, would substantially reduce the number of accidents. If landlords were also charged with the obligation of ensuring that appliances had been installed properly, there would be a further reduction. These measures are likely to be as effective as requiring an annual check and would be less expensive. Such measures might save 3 to 4 lives per year. This is a substantial proportion of this type of case but still a small number in absolute terms.

3.  Methode

3.1          Scope

Under the Gas Safety (Installation and Use) Regulations 1998, all appliances fuelled by gas, liquefied petroleum gas (LPG) etc. are within the scope of the Regulations. This includes portable LPG heaters, supplied by the landlord.

The Regulations do not cover portable room heaters supplied by the tenant, and which the tenant can be expected to take with them on departure. It is impossible, in most cases, to distinguish from records who owned a portable heater, and in any case, those owned by the landlord are already covered by the regulations. It would be unrealistic to impose an obligation on the landlord to inspect appliances if they are owned by the tenant. We therefore concentrated our attention on cases of CO poisoning from other non-gas combustible fuels. Fuels considered included coal, smokeless fuels, wood, oil or paraffin.

3.2     Statistical Data

The Office of National Statistics (ONS) was asked to supply details of accidental deaths from 1987 to 1996 where the external cause of death was carbon monoxide as defined by ICD E-Code 868 (International Classification of Cause of Death - External Cause Code).  For each case we obtained the name and address of the deceased together with date of death, age and sex, cause of death and the coroner’s district where the inquest was held.

We also obtained details of corresponding cases from the Home Accidents Deaths Database (HADD) operated by the Consumer Safety Unit (CSU) of the Department of Trade and Industry. HADD is based on data supplied by ONS but also includes information from Part II of the Coroner’s Certificate after the Inquest. Part II contains a description of the circumstances surrounding the accident but this information is not incorporated into the ONS database. From paper copies of the Part II forms, CSU adds relevant information to HADD which therefore contains more detail of cases than the ONS database alone. HADD data was only available up to the end of 1994.

There is no direct link between the two databases since the records on HADD are “anonymous” and do not include the name of the victim. However, data of death, age and sex provided us with an almost unique “key” to link the two databases. The ONS and HADD databases were loaded as separate tables into an Access database and linked to each other using this key. For the few cases where the key was duplicated in either Table and where there was no other internal information to match them, records from HADD were eliminated.

3.3          Preliminary Screening

After this process, we had 438 records from ONS with matching records from HADD for 324 of them (i.e. most cases from 1987 to 1994, but none for 1995 or 1996 where HADD data was not available). We examined all the cases in detail and eliminated those where there was sufficient evidence to indicate that the fuel was mains gas or LPG. A few other cases were identified where the cause of death was carbon monoxide from a car exhaust and/or from uncontrolled fire, usually as a result of smouldering furniture ignited by a cigarette - these cases were also excluded.

At this stage, we had 288 cases where solid fuel, wood, oil or paraffin were involved or where the fuel was unknown. The cases were spread around the country in 99 of the 147 Coroner’s Districts in England and Wales. The task then was to obtain more detail on each case by examining the files in the Coroner’s Offices.

3.4          Coroners’ Files

We contacted the Secretary of the Coroners’ Society and asked for his support. He could see no objection to what we were doing but emphasised that it was up to each coroner whether or not they chose to co-operate - they were under no obligation to do so. Most coroners, or more particularly their officers, were extremely helpful, often going to considerable lengths to assist in the work. One Coroner refused outright but did send some information on the three cases in his district. In another district, the only officer employed there could never find time to look out the files and we could get no further information.  Many officers looked up information for us in the files or photocopied and faxed documents to us. We visited a total of 79 offices obtaining information on cases in a further 20 districts by phone or fax. Of the 288 cases we had identified as requiring further investigation, we obtained additional information on 271 of them. For the other 17 cases, the files could either not be traced or we were denied access to them.

In each cases we investigated we were seeking three things:

1.  the type of fuel

2.  the status of the occupant

3.  the type of appliance and the cause of the fatality

Some files were very brief with only two or three pages. Others were much longer with detailed statements from the police, from relatives and often with a postmortem report and detailed technical reports from experts who had examined the appliance to determine the cause. Some files ran to over fifty pages.

The type of fuel was usually easy to determine. If this was found to be piped gas we did not pursue the case further.

The status of the occupant was sometimes clear though more often it was difficult to determine. In some cases there was clear reference to a landlord, usually where it was council property. In other cases there was a reference such as “when my parents bought the house the boiler was already installed”. Sometimes we made reasonable deductions - for example one victim was a “retired retail draper” and one statement was from his “cleaning lady” so we assumed he was an owner occupier. In other cases we discussed locations with the coroner or his officer and sometimes he was able to remember the case and knew the status of the victim. In other cases we obtained their comments on the area in general and made “possible” or “probable” assignments. With 50 cases where we were uncertain, we pursued local councils and Council Tax offices to see if they had additional information. We were still left with 45 cases for which we had no information on the status of the victim.  Many of the files contained detailed technical reports with information about the cause.  Where it was permitted, we took copies of these reports for later analysis. If this was not allowed, we abstracted relevant details on the spot. We obtained copies of technical files on 106 cases. Even where we had established that a case did not involve a rented property but the fuel was relevant, we collected information on the cause, where available, as causes are likely to be similar in rented and non-rented properties.

The information collected was added to the original Access database, much of it directly on to our laptop computers during the visits to coroners’ offices.

This was analysed as discussed below.

4.  The Number of Cases

4.1   Fuel Types

Details of the fuel types involved in fatal cases of poisoning by carbon monoxide from domestic heating appliances for the period 1987 to 1996 are shown in Table 1. Of the original 438 cases identified from ONS data, we were able to establish the nature of the fuel in all but 17 of the cases. 32% of all cases involve non-gas fuels, nearly all of which are solid fuels. Most solid fuel appliances use smokeless fuels such as Phurnacite. The solid fuel being used at the time of the accident is sometimes recorded but we have not analysed this information as it is the history of fuel used which could be important in accidents rather than the fuel in use at the time of the accident. As discussed later, there are some cases in which the use of the wrong type of fuel, over a period, appears to have contributed to the fatality. Wood, oil and paraffin contribute to very few cases.

LPG contributes to a quarter as many cases as non-gas fuels. The figures in Table 1 include all types of LPG fuelled appliances, including fixed ones, which, in rented accommodation, are already under the jurisdiction of the Health and Safety Executive under the previously mentioned Installation and Use Regulations. In rented property, portable appliances supplied by the landlord would also be covered by the regulations.

Further data was only collected on cases which were identified as involving a non-gas fuel.

Details on LPG cases are not considered further for the reasons discussed earlier.

4.2   Status of Properties

As discussed in Section 2, we sought to establish whether the property in which fatalities occurred was rented or owner occupied. Figures are shown in Tables 2.  Using all the information available and making reasonable deductions, we were either certain or able to make a judgement in 72% of the non-gas cases. If we exclude the cases where the fuel is not known, the proportion is 82%.

The proportion of the population living in different types of property is as follows:

Owner Occupied 67%

Council Rented 17%

Private Rented 16%

It will be noted, however, that at least 55% of the fatalities in rented accommodation are in council property. The private rentals include 2 cases rented from housing associations.  18% of cases are classified as “probably” in rented property. Most of the latter are likely to be private rentals, since council involvement is usually noted in the coroner’s files and we were unable to identify them positively through local council housing departments, however some of them will be in council properties. The balance of fatalities between council and private rental thus reflects the proportion in the total population.  61% of the non-gas cases are in owner occupied properties, a somewhat lower proportion than for the population as a whole.

Of the 141 cases known to involve non-gas fuels, 6% are in mobile homes, about double their proportion in the population as a whole.

  4.3   Cases per Year

There are some cases in which the fuel is unknown and others where rental status is unknown. We have made adjustments to allow for these “unknowns” in our estimates of the numbers of cases per year. Details are given in Tables 3 and 4.

·         Total Number per Year

Table 3 shows estimates of the number of fatal carbon monoxide poisoning cases in both rented and owner occupied properties. Column (a) shows known “non-gas” cases by year.  From Table 1 it can be seen that the 141 “non-gas” cases represent 33% of the total cases excluding the “unknown” cases. We therefore assumed that 33% of the “unknown” cases are “non-gas” ones and these are shown for each year in Column (b).  The annual totals are given in Column ©. It will be noted that the adjustments add only 4% to the overall number of cases.

·         Cases per Year in Rented Property

  Table 4 shows estimates of the number of fatal carbon monoxide poisoning cases in rented properties. Column (a) shows the number of known “non-gas” cases each year. From Table 2 it will be seen that of the 115 cases where the status is known, 38% are in rented property. We assumed that 38% of the 24 cases where the status was unknown are rented, and these are shown in Column (b). There are 17 cases which are both “unknown” status and “unknown” fuel. We assumed that 33% of them are “non-gas” (see previous paragraph) of which 38% are rented. Therefore 13% of the 17 cases are likely to be “Non-Gas” and rented. This adjustment is shown in Column ©.

The total numbers of cases in rented property are shown in Column (d). Note that the adjustments add 25% to the unadjusted figures.

  4.4   Trends

The total number of cases of carbon monoxide poisoning in rented and non-rented accommodation are compared in Table 5. There is considerable random variation from year to year and no apparent trend either in the total number of cases or in the number of cases in rented accommodation.

The results are plotted in Figure 1 on the opposite page together with lines indicating the probable trend for the total number of cases and for the cases in rented accommodation. For the overall figures, there is an apparent increase of 1.8% per year but with 90% confidence limits of ±18%. For cases in rented accommodation, there is an apparent reduction of 2.6% per year with 90% confidence limits of ±8%. In neither case are the trends significant.  In the absence of any change to regulations, the statistics do not suggest that there will be any change in the number of cases in the foreseeable future. If solid fuel appliances are replaced with gas, then there might be a reduction but there is no indication that this is happening.

4.5   Age of Victims

 Table 6 shows the age distribution of victims. There are almost no victims under 16. When all the original 438 cases are considered, there is a substantial increase in the rate with age.  This is even more pronounced with the non-gas. The pattern for rented and non-rented property is similar. The latter rates are based on the total population living in rented or non-rented property and does not allow for any differences there might be in the use of different types of heating appliances between the two sectors.

4.6   Regional Differences

Tables 7 and 8 show the regional differences in the number of deaths and in the death rates.  Deaths rate are based on the total population in each region. For the rented and non-rented sector they are adjusted for the national average split between the two sectors.

There are some considerable differences with higher overall rates in Wales and the Midlands. For non-gas cases, the rates are exceptionally high in Wales, particularly in the non-rented sector.

4.7   Monthly Variations

 Table 9 shows the monthly variation on the number of cases of poisoning by carbon monoxide from non-gas appliances. As would be expected, the number of cases is much higher in the winter with 79% of all cases between October and March. For rented properties, the seasonal variation is not quite so pronounced and the figures, based on smaller numbers, are more erratic. Nonetheless, 59% of cases occur over the winter months.

  5.  Causes

5.1   Types of Appliances

Table 10 shows the types of appliances involved in fatalities in all types of properties. For non-gas fuels, the most common appliance is the fixed room heater, followed by central heating boiler, cookers/range/stove and finally the open fire. The paraffin appliances are all portable room heaters.

Table 11 shows similar figures for fatalities in rented accommodation only. In this table we have combined the figures for the different fuels. We have also added a figure for unknown appliances so that the total matches the estimates given in Table 5.

The pattern of appliances is similar in rented and non-rented properties except that the solid fuel cooker or stove is very much a feature of the non-rented sector. Although in the past almost every household had a solid fuel range, it is only the more affluent owner occupiers who now have an “Aga”.

5.2   Factors Contributing to Fatalities

We have classified the causes of the accidents as deduced from our analysis of the coroners’ files. The number of cases in rented properties is small but causes are likely to be the same whether or not properties are rented. We have therefore included in the analysis of causes, all the cases, both rented and non-rented, for which we were able to deduce a cause. This larger base for the analysis gives greater validity to the conclusions.

The results of the analysis are given in Table 12. Often, more than one cause contributes to an accident, and up to four factors have been attributed to some cases. The Table shows the percentage of cases for which a particular factor was assigned. Because there may be several contributory causes, the total in each column is more than 100%. Figures are given separately for each type of appliance together with a total for all types. The number of cases on which each column is based, is given in the final row of the Table.

The factors have been divided into three groups. User Factors are those which are under the control of the user and about which no legislation can do anything. The only means of dealing with these areas of misuse may be to give publicity to their importance.  Installation Factors are those which should have been attended to at the time the equipment was installed. These are already covered by the Building Regulations (Part J - Heat Producing Appliances) and by British Standard Codes of Practice (BS 6461 - Installation of Chimneys and Flues for Domestic Appliance burning Solid Fuel and BS8303 - Installation of Domestic Appliances burning Solid Mineral Fuels). If these were adhered to at the time of installation then the problem should not have arisen. Maintenance Factors are factors which could be corrected by regular inspection and maintenance by a qualified person, although many of them could and should be dealt with by the user.

Before commenting on individual factors, some general comments on combustion may be useful. Hydrocarbon fuels consist mainly of molecules containing atoms of carbon and hydrogen. When these burn in a plentiful supply of oxygen the predominant products of combustion are carbon dioxide and water.

In practice, small amounts of volatile hydrocarbons (tars) and some very fine carbon particles may be formed and these can condense in the cooler walls of the chimney to produce soot. Smokeless fuels and LPG normally produce very little soot but if the supply of oxygen (i.e. air) is restricted they can do so. Bituminous coal is always liable to produce smoke and soot.

Combustion always produces small amounts of carbon monoxide, but if the air supply is restricted, levels of carbon monoxide can be much higher. Carbon monoxide is a cumulative poison which combines with haemoglobin in the blood and interferes with its ability to carry oxygen. High concentrations of carbon monoxide (2-3%) can be fatal in minutes, lower concentrations can be fatal over a period of several hours. All hydrocarbon fuels and coal products have the potential to produce carbon monoxide.

Coal also contains minerals which do not burn. Most of these remain in the grate as ash but some of the smaller particles may be carried up the flue and are deposited as fly ash when the gases cool or the flue product velocity drops. Many smokeless fuels such as phurnacite have minerals deliberately added as a binder for the briquettes.

Four conditions must be met before a fatal poisoning by carbon monoxide is likely:

1.  Air flow must be restricted so as to produce CO - this can be caused if the flue is blocked by soot, fly ash or debris, if the air inlet to the boiler is closed off or blocked by residual ash, or if there is no permanent ventilation for the room and the room is double glazed or well sealed against draughts.

2.  There must be a route for CO to escape into the room - badly fitting heater doors with worn seals, doors left open for extra heat or fuelling plugs not replaced will allow this.  However, if the flue is clear and drawing well, the flue gases will escape up the flue.  Sometimes CO can leak through a chimney into a room on the floor above the appliance. This is largely prevented if chimneys are lined, which is now standard practice when new equipment is installed.

3.  There must be insufficient ventilation in the room to disperse the CO - Building

Regulations and BS Codes of Practice require any room with a combustion appliance to have permanent ventilation sufficient to provide adequate combustion air. This was not always complied with in the past and users sometimes deliberately block the ventilation to reduce draughts.

4.  The victim must be in the room long enough to succumb - the elderly who may not move for hours at a time are particularly vulnerable but younger people who are asleep are also at risk if the other conditions are met.

The important factors we identified were as follows:

User Factors

Throat not cleared - there can be a build up of fly ash immediately above the appliance and

it may be sufficient to completely block the flow of air. Most manufacturers specify that

this area should be cleaned once a month. It is clearly the responsibility of the user and

annual maintenance would have no impact. This was the single most important factor

in fatalities. It could easily be dealt with by almost all users if they realised its

importance. Most people will remember daily routines but monthly ones are likely to be need to be aware that if their appliance stops drawing well some attention is needed.

They will have to have help and it should be the responsibility of the landlord to clear the flue but he or she needs to be told of the problem.

Doors open - some users leave the door of the heater open for extra warmth. This is important for fixed room heaters and publicity is needed to remind people that it is an unsafe practice.

Ventilation blocked - some users deliberately seal permanent ventilation to reduce draughts. This is particularly important with portable heaters where there is no flue to carry away combustion products.

Ash not cleared - if the firebed is not riddled and ash removed regularly (daily) air supply will be restricted. Most users should be able to deal with this but it may be difficult for some of the elderly.

Unsuitable fuel - the use of bituminous coals or household rubbish on appliances designed for smokeless fuels produces excess soot and may block the flue. This is only an occasional problem.

Other misuse - miscellaneous factors such as failing to replace the refuelling plugs or leaving access points on the flue open.

Other medical factors - some users have medical conditions, such as heart disease,

which make them particularly vulnerable to even small levels of CO

Installation Factors

Inadequate ventilation - failure to install the required amount of permanent ventilation.  This can be exacerbated if the room is later double glazed or if draught excluders are fitted. This is an important factor and one which landlords could be required to check, on a one-off basis. It is particularly important with portable heaters without a flue but of course unenforceable since such heaters are never “installed”.

Poor cleaning access - this means that flues cannot be adequately or easily cleaned. Codes of practice require easy access to all parts of the flue. Landlords could be required to bring flue design up to scratch, again on a one-off basis.

Unlined chimney - unlined chimneys are subject to acid attack and can begin to leak flue gases into other rooms. Unlined chimneys also cool the flue gases more quickly so that the fire draws less well and is more likely to allow CO to leak back into the room. Codes of practice now require all flues to be lined but this may not have been done with many older installations. Landlords could be required to install lining if it is missing.

Terminal badly sited - terminals should be placed so that they are never in an area of positive pressure, whatever the wind speed or direction. In some circumstances, a down draught can be generated, blowing combustion products back into the room.

overlooked. Is it possible to design the throat area so that it is less prone to blockage?

Flue blocked by debris - this may be fallen masonry or mortar or even a bird’s nest. It may happen at any time and is likely to lead to a sudden deterioration in performance. It needs dealing with when it occurs.

Flue badly designed - this is frequently a factor which affects access to the flue and subsequently makes cleaning it difficult.

Maintenance Factors

Flue not swept - if the flue is not swept regularly it can gradually become blocked with soot.  It is recommended that flues are swept annually. In most council property it is the tenant’s responsibility to have the chimney swept.

Leaks from doors etc - door and other seals need replacing from time to time.

Faulty appliance - faults in appliances are rare.

Lack of maintenance - may refer to any of the above three points.

Poor maintenance - occasionally maintenance creates the problem, when for example, covers are not properly replaced.

5.3   Action

As the cost of introducing specific regulation controls for rented accommodation is likely to be very high, thereby placing burdens on business and increased rents for consumers, and given that considerably more deaths occur in the owner occupied housing sector, a more effective response is to raise public awareness through publicity. This would not only be more cost effective, but would reach households in both the rented and non-rented housing sectors.

To this end, the Department of the Environment, Transport and the Regions have produced a safety advice leaflet specifically to warn of the dangers of non-gas fuels. Coupled with this, during the autumn and winter of 1997/98, the DTI was active in alerting consumers to the dangers of CO by enclosing leaflets in domestic gas bills and distributing leaflets to doctors surgeries throughout the country and continues to underpin this with TV adverts. All DTI material refers to all fuels.

Related Links:
Carbon Monoxide Lawyers

Related Searches:
Carbon Monoxide Deaths
Carbon Monoxide Suicide