How widespread is the groundwater arsenic problem in
Bangladesh?
It is very widespread. We carried out surveys in 41 of the 64 districts
and compiled the existing survey data for all of the country. Our
survey area included what were thought to the worst affected areas
of southern and eastern Bangladesh. Here, some 35% of the wells tested
were above the Bangladesh limit of 0.05 mg/l and 51% were above the
WHO Guideline value of 0.01 mg/l. Combining our data with all the
other surveys it seems that about 26% of all the wells in the country
might be affected. However, different surveys were done in different
ways and so this figure should be taken as a rough guide rather than
a precise calculation.
Where is the problem worst?
The extent of contamination varies greatly across the country. Both
our data and most of the earlier surveys show that the most seriously
affected areas are the south-east of Bangladesh around the Meghna
Estuary and some parts of the south-west. The worst affected districts
are Chandpur, Noakhali and Lakshmipur in the south-east and Madaripur
in the south-west. Apart from these general patterns, it needs
to be appreciated that there can be pockets of very severe contamination,
referred to as
‘hot-spots’ in districts with little overall contamination.
Are there any arsenic-free areas?
That’s a difficult question to give a simple answer to. Although
three quarters of the wells in the country are estimated to conform
to the Bangladesh standard, no district has been tested to be 100%
arsenic-free. Nevertheless, the percentages of wells contaminated
vary from nearly 90% in Chandpur to as low as 0.3% in Rangpur, Thakurgaon
and Panchgarh in the far northwest. It’s also true that the
distribution of arsenic varies according to the geological conditions.
There are older aquifers beneath the Barind and Madhupur Tracts that
are virtually free of arsenic while the adjacent floodplains may
be badly affected. The little bit of evidence we have suggests that
the hilly areas also have very little contamination, but there are
not enough test results to be really confident about that at this
stage. As well as varying from area to area, the arsenic concentrations
can also change as you go deeper below the ground surface, and most
wells deeper than 150 - 200 metres (500 – 650 feet) are also
safe.
Your survey shows that the Nawabganj district in western
Bangladesh is generally low in arsenic, but Chapai Nawabganj
is a well-known arsenic ‘hot spot’, where high arsenic
concentrations in groundwater in Bangladesh were first discovered.
How do you explain that?
Yes, we were surprised by that too as we had already made a small
study of Chapai Nawabganj town and knew it to be very high in arsenic.
However, it appears to be an isolated hot spot in a generally low
arsenic area and it was simply too small to be 'hit' by our survey
which only took about eight samples per thana. Using random sampling,
we would have to take many more samples - perhaps four times as many
- to identify all such hotspots. Something we have found is that
this ‘patchiness’
is characteristic of certain parts of the country, particularly along
the Ganges valley. By comparison, the contamination in other areas
like around the Meghna Estuary is much more uniform.
Is Dhaka City affected?
No. Dhaka City takes its water from an older aquifer (the Dupi Tila
sands) and many tests have shown that it is essentially arsenic-free.
A few of these tests even go back ten years. So there is no problem
in the city but just south of the city, in the rural parts of Dhaka
district, there is a problem. Here the water is derived from younger
aquifers which is where the problem occurs. You have to be careful
when talking about Dhaka to distinguish the city from the district.
How many people are affected?
Groundwater provides the drinking water for more than 90% of people
in Bangladesh. Using the estimated 1998 population, we estimate
that around 20-30 million people are drinking water with arsenic
above the Bangladesh standard. However, some areas have been better
surveyed than others and this estimate may change slightly as more
results come in.
If the Bangladesh standard for arsenic in drinking water
were lowered from 0.05 mg/l to the WHO Guideline Value of 0.01
mg/l, how many more people would be affected?
In our survey area, we found many wells with below 0.01 mg/l arsenic
and quite a few above 0.05 mg/l, with relatively few in between.
However, the results of other surveys suggest that this proportion
may be higher in other areas. We estimate that if the Bangladesh
standard were changed to 0.01 mg/l, the water supplies for perhaps
another 10 million people would fail to meet the standard.
Is it getting worse?
There is a tendency for people to hear reports of arsenic being found
in areas that had not been tested before and mistakenly think that
the problem is spreading. There are still a few areas that have
not been tested at all. The other question is whether the arsenic
contents of wells are increasing. Nobody has been monitoring wells
for more than a few years so nobody can say with absolute certainty
whether conditions are really changing, although a few wells have
increased in concentration. We do know that more of the older wells
are contaminated than those constructed in the last couple of years,
so it might be getting worse, but it will take a few years before
anyone can answer this for sure.
An early suggestion was that the arsenic might be coming
from the wood preservative used in the rural electrification
schemes. Is that possible?
No, definitely not. The quantity of arsenic found in the groundwater
is much too large for that and arsenic-affected wells are found in
areas where there are no rural electrification schemes. Sometimes
arsenic appears to increase with the depth of well which is the opposite
of what you would expect if the arsenic were leaching from the soil.
A hypothesis put forward by West Bengali scientists
said that the arsenic is due to excessive pumping of groundwater
for irrigation over the last 15 years or so. The hypothesis says
that the increased seasonal lowering of the water table has promoted
the oxidation of the arsenic-containing sulphide mineral, pyrite,
leading to the release of arsenic to the groundwater. Is this
true?
We don’t think so. The evidence from the field just doesn’t
seem to fit the theory. If you look at a map of the areas where the
arsenic contamination is worst and compare it with the places where
the water table is deepest in the dry season or where most groundwater
is pumped, you don’t find any connection between them. Also,
if pyrite oxidation were the key to the arsenic problem, we would
expect to find quite high concentrations of sulphate in the groundwater.
We don’t. So we don’t find any support for the idea that
excessive pumping has created the arsenic problem in Bangladesh.
In other parts of the world phosphate fertilisers have
been blamed for mobilising arsenic into groundwater. Do you think
they might be to blame here?
No. It is true that many Bangladesh groundwaters have high concentrations
of phosphate and we know that phosphate may make the arsenic more
soluble. Some scientists believe that these high phosphate concentrations
have come from the use of phosphate fertilisers but we believe that
the most likely source is from the sediment itself, just like the
arsenic. This matter will be resolved as we learn more about the
movement of phosphate and arsenic in Bangladesh soils and aquifers.
So where do you think the arsenic is coming from?
The arsenic is coming from the alluvial sediments that make up much
of Bangladesh. It was deposited together with the sediments, mostly
in the last 20 000 years or so. Exactly how and when the arsenic
dissolved into the groundwater is still uncertain but it is likely
that it was originally derived from the iron and maybe manganese
oxides that are found in the sediments. It only takes a little
arsenic to dissolve from the sediment to give a serious contamination
problem.
So why has the problem only just appeared?
There are anecdotes dating back more than 20 years of various skin
complaints being a problem even then. The main reason, however,
is that most of the wells have only been put down in the last 20
years or so, and the symptoms of chronic arsenic poisoning may
take 10 to 15 years to reveal themselves. Before that, surface
water and groundwater from shallow dug wells was used and we now
know that these sources both tend to be low in arsenic.
Is it still being released from the sediments?
It is difficult to say at the moment but it probably is in the most
recent sediments in the those parts where the delta is still pushing
out into the Bay of Bengal, like the south of Noakhali. On the
other hand, conditions may have more or less stabilised further
inland. The chances are that, in most areas, the arsenic has been
in the groundwater for thousands of years and is now just being
pumped out. We need to monitor some shallow and deep wells very
carefully over a number of years to prove whether it is changing
with time.
Can you use your knowledge of the geology of Bangladesh
to predict which areas will be high in arsenic and which will
be low?
We know that the problem is largely confined to the alluvial sediments
of Holocene age. We also know that wells in the older and deeper
aquifers are largely arsenic-free and that this is related to geology.
Patterns are emerging that relate the occurrence of arsenic to the
type of sediments and how they were deposited, but there is still
a lot of variation in these geological units that we cannot explain
in detail.
Your survey only looked at part of Bangladesh. Why did
you do that?
That was a difficult decision to make. We wanted to be as useful
as possible as quickly as possible and so when planning the survey
in mid-1997, we decided to concentrate on the areas that were then
thought to be worst affected by the arsenic. In fact, we added some
districts in the north-east region later as it was becoming clear
that parts of that area were also affected.
Are you going to survey the rest of Bangladesh?
Yes, we have a plan to survey the northern part of Bangladesh but
the details have not yet been finalised. There is no doubt that
there will have to be a massive testing programme to identify all
of the affected wells.
It appears that people are showing symptoms of arsenic
poisoning in some of the worst-affected areas of Bangladesh (as
shown by your map) and not in others. Can you explain that?
That is very interesting but we do not know the reason why that is
for sure. There seem to be a lot of arsenic patients from the Pabna
region in the west but relatively few from the Chandpur region in
the east, one of the worst-affected regions according to our map.
This could be due to differences in the iron content of the water
or perhaps due to differences in diet but it also may reflect the
geographic nature of the arsenic distribution and the scale of the
exceedances. In the west of Bangladesh, the arsenic distribution
appears to be quite patchy with the occasional village with a localised
cluster of wells with exceptionally high arsenic concentrations,
1 mg/l or more. These hot spots give rise to a relatively large number
of arsenic patients – indeed, arsenic-affected patients are
one of the best ways of locating such hot spots. They are difficult
to locate reliably with randomised sampling when the sample density
is only one sample per 40 km2 or so, equivalent to about
one well in a thousand, as in our regional survey. In the east, while
the percentage of contaminated wells is higher on a thana or district
basis, the extremely high concentrations appear to be less common.
Therefore it may take longer before symptoms appear. The total number
of arsenic-affected patients so far identified, a few thousand, is
only a very small percentage of those exposed to concentrations above
the Bangladesh or WHO standards and so the pattern of distribution
of patients may change as the exposure period lengthens.
Why are the results of your survey different from those
of other surveys by DPHE, NIPSOM and Dhaka Community Hospital?
Actually, we don’t think the results are all that different.
Insofar as they are different the main reason is because of the way
the wells were selected. Our survey aimed at getting an overview
of the distribution of arsenic in the worst-affected areas to provide
reliable statistics of the number of wells affected and to identify
any geological patterns. We never knew whether a well was high in
arsenic or low when sampling it. Other surveys had different objectives
and so chose their wells differently. For example, many surveys only
sampled wells where it was suspected there were arsenic-affected
patients. This will obviously give different statistics from a survey
where wells were selected randomly. Both types of survey are valid
in their own way, but it must be appreciated that the way you select
the wells will determine the answers you get.
But if the ‘hot spots’ are so small, how are
we going to find them all?
The first thing to do is to follow up all villages where symptoms
of arsenicosis are found with an arsenic-testing programme. Dhaka
Community Hospital is already doing this under the Rapid Action Programme.
Two hundred villages have been studied in detail and another 300
are on the list to be studied. The next thing to do is to continue
with a systematic testing programme to identify less badly affected
areas. It will be important to co-ordinate the data collection so
that wells are not sampled twice and so that the results of the tests
already carried out are not lost.
Do you think that the hot spots will spread with time?
Not very much, or at least not over a few decades. The groundwater
moves very slowly because the land of Bangladesh is so flat and
so hydraulic gradients are very low. It is only near rivers and
irrigation wells that significant flow of groundwater is induced
and even here, the arsenic will only move rather slowly. Below
10 or 20 metres there is not much movement to spread the hot spot
laterally. In any case, the movement of arsenic is slower than
that of the water because some of it is adsorbed by the sediments.
Would it be possible to pump all of the arsenic away
then?
No, at least not in a matter of a few years. The quantities of water
involved are enormous, and sometimes the arsenic is deep in the aquifer.
Also the arsenic may be continuously replenished from the sediments.
On the other hand, active groundwater circulation near the surface
is greater than at depth and it may be that the shallow zone could
gradually become depleted in arsenic by natural flushing. In fact,
our data suggest that this may have already happened to some extent.
Do you think that it is feasible to test all the wells
in Bangladesh for arsenic?
That would be an enormous task because of the large number of wells
involved – perhaps four million altogether - but it could, and
probably must, be done some time. Clearly, the priority is to test
the worst-affected areas first and to follow this up by providing
an arsenic-free drinking water supply for the people using wells
that are contaminated.
How would you do that?
The testing will undoubtedly involve a lot of field testing as well
as laboratory testing and will have to be run a bit like a military
campaign. We really need better field test kits – especially
those able to get more reliable data in the 0.04-0.06 mg/l
region. There are a number of such kits under development at the
moment but even the present kits are good enough to identify the
worst-affected wells.
I’ve heard it said that a well tested ‘no’
for arsenic one day and ‘yes’ another day. Is that possible?
Yes, it’s possible, though it should be quite rare. There are
two reasons why this could be true. Firstly, there may be some genuine
reason why the water is changing in quality
– a change in the flow path due to seasonal variations, for
example. The other reason is that no test kit is perfect and so if
the water is close to the 'yes-no' cut-off, it could be 'yes' one
day and 'no' the next. This is one of the problems with ‘yes-no’ kits.
Even laboratory analyses are subject to some analytical uncertainty
and this always has to be remembered. We are trying to measure very
small quantities of arsenic and this is not easy to do.
Can we trust your results?
Yes, we hope so! We took a lot of care to ensure the best quality
data possible. For example, we tested some samples in two BGS laboratories
and the results agreed well. We were also careful in sampling.
The range of arsenic concentrations in Bangladesh groundwaters
is very large – from less than some 0.0005 mg/l to more than
2 mg/l so it is not always necessary to have high precision analyses
to separate contaminated from non-contaminated groundwaters. But
in general it does help of course. Obtaining reliable statistics
for the percentage of wells contaminated in each district is much
more challenging and it remains to be seen how well our current
estimates stand up in the future.
It is often said that deep wells are free of arsenic.
Is that true?
Yes. We found very little arsenic contamination at depths greater
than 150-200 m or approximately 500-650 ft. Only about 1% of the
deep wells we tested were contaminated. Even the earlier surveys,
where we suspect the well depths were not always recorded accurately,
showed that only 2-3% of the deep wells were contaminated. The main
reason for the deep wells being safe is probably because the sediments
are older and any arsenic present was flushed out thousands of years
ago. Nevertheless, it is recommended that all new wells be tested
before being put into use.
Results show that some of the new deep wells that were
arsenic-free became contaminated with arsenic over a couple of
years of use. How do you explain that?
We have heard anecdotal reports of this happening, mostly from West
Bengal. If it is happening within a few years, then it’s most
likely due to short-circuiting down the outside of the tubewell pipes.
We don’t think that a big change in the aquifer could occur
in such a short time. This reminds us how important it is to provide
a proper sanitary seal for all tubewells, irrespective of whether
arsenic is suspected.
Does irrigation have any effect on the arsenic problem?
Another difficult question! We are fairly sure that irrigation is
not the primary cause of the problem but that does not mean it
has no influence at all. The distribution of arsenic is not associated
with the lowering of the water table or the amount of pumping,
however, the use of phosphate fertilisers might have an effect
on arsenic in the long run. We suspect that a long-term influence
of irrigation could be its impact on the aeration status of the
aquifer since paddy fields are rather efficient ways of reducing
the input of air to the underlying aquifers. Then there are the
questions of whether arsenic in contaminated irrigation water is
getting into human or animal food chains. Preliminary data suggest
that arsenic is not accumulating in rice grain, but these questions
need thorough investigation.
Is the arsenic in the water related to the iron content
of the water?
Yes and no. By international standards, the iron content of most
Bangladesh groundwaters is high, which reflects the reducing nature
of the aquifers. There is a relationship between iron and arsenic
in the water, but it is slightly complicated. Almost all of the waters
that contain high arsenic also contain high iron. However, there
are many waters containing lots of iron that contain little or no
arsenic, and a few waters with high arsenic and no iron. So, unfortunately
you can’t really use the iron content to predict the arsenic
content.
Is there anything we can do to reduce the arsenic in
our drinking water at village level?
Yes. A lot of people are working on that. The World Bank recently
announced an initial loan of $44 million to the Government of Bangladesh
to look at various mitigation options and begin implementing them
on an emergency basis. It is likely that different solutions will
be found for different situations ? for example, it depends whether
most of the wells in a village are affected or not. In some places,
the naturally high concentration of iron in the water will help.
If the water is allowed to stand overnight and the iron sludge allowed
to settle, then the arsenic concentration in the clear water above
will be reduced. This is already done in some places and while it
will not completely solve the problem, it is easy to do, costs little
and could help to reduce the arsenic intake immediately. Another
simple approach that needs further testing is to use alum to coprecipitate
some of the arsenic – alum is already used in Bangladesh to
clarify water after a flood and the same approach could probably
be adapted to help reduce the arsenic intake. Rainwater could also
be used for at least some of the year. All these ideas need testing
for acceptability and reliability at village level and if successful,
will require a massive publicity campaign to promote them.
Should we return to using surface water like we used
to do?
It’s basically a question of balancing the risks between the
arsenic and diarrhoeal disease. Clearly the arsenic problem in groundwater
must be solved as a matter of urgency. It is generally agreed that
there is no single solution of universal applicability. Surface water
with treatment will, no doubt, play a role in the mitigation strategy.
There are two main problems with using surface water – first
to make sure that the treatment is done properly and routinely, and
second that in many areas there is a serious shortage of surface
water in the dry season. Neither of these problems is insurmountable,
but experience shows that they are not trivial either, and in some
cases may also be expensive and slow to implement.
Would you seal all high arsenic wells?
Sealing contaminated wells is controversial. The critical point is
not to drink arsenic-contaminated water. Painting wells green and
red is a good idea since it highlights which wells have been tested
and whether they are contaminated or not. Red wells are fine for
washing and so could save a lot of walking. But some wells have
so much arsenic in them that it might be better to leave the water
in the ground rather than to even use it for washing or to attempt
to treat it. It therefore seems a reasonable idea to seal these
wells, at least temporarily.
Are the filters that you can buy any good for removing
arsenic?
It is possible in principle to filter out much of the arsenic using
filters of various designs - the importance of iron precipitation
in reducing arsenic in water has already been mentioned and this
principle can be used to advantage in filters. A number of groups
are working on filter designs and it remains to be seen whether they
can provide an effective and acceptable long-term solution.
Do you think that there are other places in the world
that have a similar arsenic problem to Bangladesh but don’t
yet know about it?
Almost certainly, but we suspect that the scale of the problem in
Bangladesh is exceptional ? after all, it is part of one of the largest
deltas in the world. The lesson to be learnt from Bangladesh is that
arsenic must be on the list of chemicals tested for when assessing
the quality of water for drinking.
Are there likely to be other trace element problems
with the groundwater in Bangladesh?
We have not looked at this exhaustively but expect that any other
problems will be much less severe in comparison to arsenic. Concentrations
of manganese and boron exceeding the WHO health-related guideline
values have been found in a significant proportion of wells. A very
small number of samples also exceeded the WHO health-related guideline
values for barium and chromium. The limited published data for fluoride
in Bangladesh suggest that concentrations are low. Concentrations
of ammonium also exceed the WHO guideline value in some wells. In
some areas of Bangladesh, there is serious iodine deficiency in the
diet, which is exacerbated by the low iodine content of the groundwater.
Elsewhere iodine deficiency has been tackled by using iodised salt.
The nitrate concentration is very low in most groundwaters but it
can be high in some contaminated shallow groundwaters. Similarly,
there is some risk of pesticides leaching from agricultural fields
to the water table and eventually to water supplies. However, in
general and apart from arsenic, microbiological contamination poses
the greatest cause for concern.
Iron removal by aeration and sedimentation at Faridpur
