ADDENDUM 3—REPORT ON SUPPLY/DEMAND
SENT TO DTI IN MARCH 2005
ANALYSIS OF UK GAS SUPPLY/DEMAND FOR DTI
ENERGY TEAM
SUMMARY
This paper looks at the supply/demand balance
for next winter under several scenarios, including the loss of
Interconnector import. The analysis is based on real data of actual
supply, demand and temperatures seen and reported publicly for
the UK.
Recent experience has highlighted vital flaws
in the assumptions underpinning all previous UK gas supply/demand
models. These are:
(i) That the Interconnector will import
at maximum rates when demand and price are high in the UK market
and
(ii) The level of deliverability of beach
gas.
Supply Capability
The supply capability of UK gas fields is reducing
at an alarming rate. The reduction in supply capability has already
exceeded the capacity of the Interconnector and the main UK storage
facility at Rough. The capacity of UK gas fields is well short
of winter demand requirements. The UK is critically dependent
on storage, imports from Norway and Interconnector imports. For
the past few winters the combination of supplies has been sufficient
to meet demand but this is mainly due to having very mild winters
and a higher capacity in the UK gas fields.
The forecasts made by Transco for beach supplies
to the UK seem to consistently over state the amount of gas available
to meet UK demand. For winter '04/05 Transco assumed peak supplies
of 364 mcm/day and sustained supplies of 346 mcm/day. In practice
we have seen a peak of 331 and an average of 305 mcm/day from
November to February. This "shortfall" of 30-40 mcm/day
is more than the current Interconnector import capacity.
Next Winter
Looking forward to next winter there are two
key projects increasing the supply available to the UK. One is
the new LNG facility at the Isle of Grain and the other is increased
import capacity on the Interconnector. If both of these projects
are delivered, the maximum supplies of gas available to the UK
should be able to cope with a "one in three" winter
demand but are not able to cope with winter demands that can be
expected to occur one in every six winters.
If there is a delay to the Interconnector project
or gas is not available to fill the new capacity then supplies
are not sufficient for demands that can be expected in a "one
in three" winter.
If the UK experiences the combination of cold
weather and lack of supply from the Interconnector (as seen in
March 2005), it is expected that significant shutdowns of the
UK customer base will be required within two weeks. The combination
of loss of Interconnector supplies and cold weather should not
be considered as independent events. Instead, they are most likely
to be coincident as cold weather in the UK is likely to coincide
with cold temperatures on the continent increasing the likelihood
in loss of gas imports.
SCOPE OF
ANALYSIS
This analysis was prompted by the experience
during the recent three-week cold spell in the UK, which saw a
combination of high UK demand and low Interconnector flow. This
short cold spell saw key UK storage stocks fall to alarmingly
low levels from a healthy position and prompted concerns about
supply capability for next winter. This paper attempts to assess
the supply/demand balance in the UK and assess whether the UK
is likely to run short of gas next winter. Previous models of
UK supply have assumed that high UK prices relative to continental
wholesale prices would lead to maximum imports to the UK during
high demand.
During the recent cold spell UK prices reached
£1.60/therm while imports through the Interconnector were
close to zero. The average Day Ahead price over the 18-day period
from 21 February to 10 March was around 65 pence per therm. This
compares to a wholesale price of gas in Continental Europe is
about 28 pence/therm at present.
ANALYSIS METHOD
The analysis is done in six stages:
| Stage 1: | Define the relationship between temperature and gas demand for UK market.
|
| Stage 2: | Using data from 1970 to 2004 assess the probability of various temperature events.
|
| Stage 3: | Use the temperature data and the correlation from stage 1 to define likely UK gas demand scenarios.
|
| Stage 4: | Assessment of UK gas field production capacity for next winter.
|
| Stage 5: | Assess total supply capability for next winter.
|
| Stage 6: | Compare supply capability to demand requirements for next winter.
|
| |
ANALYSIS RESULTS
Stage 1: Define the relationship between temperature and
gas demand
The average monthly temperature data was taken from DTI table
1.1.8. The "Gas Output from the Transmission System"
from DTI table 4.2 was used as a basis for demand.
Combining these two sets of data gives the graph on the following
page. It is easy to see that there is a very strong correlation
between temperature and demand. Statistically there is a 98% correlation
between the two. From this data it is possible to determine a
formula (given in the graph below) linking the two sets of data.
This includes the demand for the UK but does not include the use
in the transmission system (about 2 mcm/day) or the use in Ireland
(about 13 mcm/day).
Including the demand from Ireland and the use in the transmission
system the formula used in this analysis is:
Gas Demand (mcm/day) = 0.525 * (temp C)2—28.8
* (temp °C) + 508
This gas demand is made up of two basic categories. There
is a stable demand of around 80 mcm/day consumed by sites directly
connected to the NTS. This is a combination of power stations
and large industrial demand and does not vary with temperature.
The remaining demand is delivered via the LDZ's (Local Distribution
Zones). This demand is a combination of domestic and small industrial/commercial
customers. The LDZ demand is highly temperature dependant.
Stage 2: Assess the probability of various temperature
events
The temperature data from DTI table 1.1.8 gives monthly average
temperatures for each month from 1970 to 2004. Using this data
it is possible to estimate the likelihood of various winter scenarios.
Rather than look at rare events like a "one in 50" winter,
we have used the data to assess scenarios that are quite likely.
The table below shows the results of this analysis.
| | |
| | | |
| 1970-2004 | Below 5C
| Below 4C | Below 3C
| Below 2C | Below 1C
| Below 0C |
| | |
| | |
|
One Month | 29 |
16 | 11 | 6 | 3
| 1 |
| Two Months | 20 | 11
| 3 | 1 | 0 |
0 |
| Three Months | 8 | 0
| 0 | 0 | 0 |
0 |
| One Month | 83% | 46%
| 31% | 17% | 9%
| 3% |
| Two Months | 57% | 31%
| 9% | 3% | 0% |
0% |
| Three Months | 23% | 0%
| 0% | 0% | 0% |
0% |
| | |
| | | |
| | |
| | | |
In the period from 1970 to 2004 there are 35 winters. Of
these, 29 have seen at least one month below 5°C, 20 have
seen two months below 5°C and 8 have seen three months below
5°C. This can be converted into a probability; ie there is
an 83% chance that a winter will have at least one month with
temperatures below 5°C, a 57% chance of two months below
5°C etc. The blocks in yellow in the table above give the
key probabilities used in this analysis. In summary for next winter
the key probabilities are:
There is a 9% chance (1 in 11) that we will experience one
month below 1°C.
There is a 17% chance (1 in 6) that we will experience one
month below 2°C.
There is a 31% chance (1 in 3) that we will experience one
month below 3°C.
There is a 46% chance (1 in 2) that we will experience one
month below 4°C.
There is a 31% chance (1 in 3) that we will experience two
months below 4°C.
These are relatively high probability events. It is easy
to become complacent about the chance of cold weather because
we have experienced several of the warmest winters on record over
the past few winters. The chart below shows the lowest winter
temperatures each year since 1970. The period in the mid 70's
saw a dramatic change from "warm" to "cold"
winters, demonstrating that it is entirely possible to see a sudden
return to "cold" conditions.
Stage 3: Develop a demand profile for winter 2005-06
From the correlation from Stage 1 and the temperature profiles
from Stage 2 it is possible to determine the likely demand requirements
for next winter. The table below shows the demand expected for
each temperature.
| | |
| | | |
| 1970-2004 | Below 5C
| Below 4C | Below 3C
| Below 2C | Below 1C
| Below 0C |
| | |
| | | |
One Month | 83% |
46% | 31% | 17% |
9% | 3% |
| Two Months | 57% | 31%
| 9% | 3% | 0% |
0% |
| Three Months | 23% | 0%
| 0% | 0% | 0% |
0% |
| | |
| | | |
| Demand (mcm/day) | 377 | 401
| 426 | 453 | 480
| 508 |
| | |
| | | |
| | |
| | |
|
This demand can be converted to graphical format giving a
demand/probability profile as shown in Appendix 1.
Stage 4: Assessment of UK gas field production capacity
for next winter
It is well recognised (Transco Winter Outlook Report and
DTI commentary on gas supply figures) that the UK production of
natural gas has been declining rapidly in recent years. The following
graph shows the DTI data for UK gas production in each of the
last nine winter periods. The graph shows the best month and the
best "two-month" period for each winter. At the current
rate of decline the UK sources will only be able to produce at
a monthly average rate of 275 mcm/day by next winter. The reduction
in capacity that has occurred since 1990 is more than the combined
capacity of the Interconnector and long-range storage.
Stage 5: Assess total supply capability for next winter
The Supply capacity available to the UK comprises:
(i) UK gas production (described in Section 3);
(ii) Imports from Norway (this gas can flow to the UK
or directly to Continental Europe);
(iii) Imports through the Interconnector (ICUK); and
(iv) Various storage facilities (generally filled in summer
and emptied out during winter).
When added to the other supply sources it is possible to
define a "supply stack" for the winter. During 2005
there should be a new LNG import facility coming on line (Isle
of Grain conversion project) and by the end of December 2005 the
import capacity of the Interconnector should be increased by the
installation of compressors at Zeebrugge. From these different
supply options we have developed four supply scenarios as shown
in the table on the next page.
| | |
| | |
Modelling Assumptions UK Gas Fields and Imports
| | Scenario 1 High Import
| Scenario 2 Current Import | Scenario 3 Low Import
| Scenario 4 Export |
| | |
| | |
UK Gas Fields | mcm/day
| 274 | 274 | 274
| 274 |
| Norwegian Imports | mcm/day |
40 | 37 | 28 | 17
|
| Interconnector | mcm/day |
47 | 25 | 0 | -50
|
| Isle of Grain LNG | mcm/day |
10 | 10 | 0 | 0
|
| | |
| | |
Beach | mcm/day |
314 | 311 | 302 |
291 |
| | |
| | |
| | |
| | |
Scenario 1: Maximum supply from all supply sources.
Scenario 2: Current import from Norway and ICUK, with
new import from IOG.
Scenario 3: No imports from ICUK or IOG. Norway import
at levels during winter 2003-04.
Scenario 4: Interconnector in export mode. Norway import
at levels during winter 2002-03.
The expected "Beach" supplies (UK production and
Norwegian Imports) for each scenario are given at the bottom of
the table. The beach supplies for the current winter (2004-05)
averaged 305 mcm/day. If the beach supplies are declining at 7%
per year (as reported by Transco and DTI) the expected beach supplies
would be 284 mcm/day by winter 2005-06. All of the scenarios assume
beach supplies that are higher than this.
Assuming all of the supply sources at maximum capacity the "supply
stack" looks like:
This shows the maximum capacity that is available by next
winter. By using ALL of the storage facilities at maximum rate
and assuming maximum imports from Norway and the Interconnector
it should be possible to supply demand of 488 mcm/day for five
days. After five days at this rate LNG storage would be empty
and the maximum capacity falls to about 435 mcm/day. After only
18 days most of the Medium Range storage facilities would be empty
and supply would fall to a level of about 418 mcm/day.
This "supply stack" depends to a very large extent
on assumptions about Norwegian imports and the Interconnector
flows and to a lesser extent on the availability of LNG at the
new Isle of Grain terminal. During the recent cold spell we saw
the gas Interconnector reduce to virtually zero flow as a result
of cold weather on the continent. It is reasonable to assume cold
weather in the UK is likely to be associated with cold weather
on the continent and therefore gas is less available from the
continental sources at times when the demand for it is most urgent.
This is the basis for Scenario 3, which assumes a reduction in
Norwegian imports and zero imports from the Interconnector.
The graph below shows the four supply scenarios.
Stage 6: Compare supply capability to demand requirements
for next winter
The graphs in Annex 2 combine the supply and demand figures
from the previous analysis. They are shown with the best supply
option first.
Scenario 1: Maximum Supply
With maximum supply it is possible to meet the demand for
any 3°C months but difficult if the temperature is 2°C
or below. For months below 2°C there would need to be significant
demand side response to reduce demand. There is a 17% chance (one
in six) of demand response being required even with all new supplies
available at maximum capacity.
Scenario 2: Current Imports
The maximum supply case assumes the increased capacity through
the Interconnector is available. This is scheduled to be on line
at the end of December 2005. If this increase is not available,
the supply picture is shown by the second graph in Annex 1. In
this case the UK struggles to meet demand if the temperature is
below 3°C. There is a 31% chance (one in three) of experiencing
a month like this in the coming winter. Under this scenario very
significant Site closures to reduce UK gas demand could be required.
As can be seen from the graph the supply shortfall is about 80
mcm/day if the UK has a month below 1°C. This is the capacity
of all the direct NTS demand which includes power stations.
Scenario 3: Low imports
During the very recent cold spell the UK experienced a combination
of events that give great cause for concern. The cold weather
in the UK (temperatures were below seasonal normal but were not
exceptionally low) was coincident with cold weather on the continent
and supplies of gas through the Interconnector effectively stopped.
Additionally it is understood LNG ships were not available to
supply Zeebrugge. This set of real events is the basis of the
third scenario. This is shown in the third graph in Annex 1. As
can be seen from the graph this level of supply causes severe
problems even under the mild 5°C demand conditions. This
supply scenario would, very quickly, lead to closure of large
parts of the UK supply system including parts of the Local Distribution
Zones. Under this scenario there may not be enough gas to preserve
supplies for domestic customers.
Scenario 4: Interconnector Exporting
The fourth graph shows the impact of the Interconnector exporting
during winter months. There is not enough UK supply to even begin
to meet demand in this scenario. The UK would have to shut the
Interconnector in this scenario to avoid a possible National Emergency.
This would effectively convert this scenario to the same as scenario
3.
CONCLUSIONS
This analysis shows that events we have already experienced
during the current winter would lead to dramatic gas shortages
during next winter. In recent years the UK has not faced severe
shortages because we have had more supplies from UK gas fields
and we have been fortunate to experience some of the warmest winters
on record.
There are two key core assumptions in this analysis that
differ from any analysis we have seen previously. They are:
(i) Beach supply capability—in winter
2004-05 this has been much lower than assumed in the "Winter
Outlook Report" published by Transco in October 2005. For
winter 2004-05 Transco assumed peak supplies of 364 mcm/day and
sustained supplies of 346 mcm/day. In practice we have seen a
peak of 331 and an average of 305 mcm/day. This "shortfall"
of 30-40 mcm/day is more than the current Interconnector import
capacity. We consider that the beach flows reported do represent
a good measure of the capability of the off-shore infrastructure
given the (high) level of NBP prices in the prompt market. We
question why there is such a large difference between the expected
and actual beach flows. Our expectation is that there will be
further decline for the coming winter.
(ii) Interconnector behaviour—it has been assumed
that in winter the Interconnector will provide a base load flow
up to its physical flow capacity. This winter, we have seen very
real evidence that this assumption is not robust, especially at
times of cold weather, when demand is highest.
Looking forward to next winter, the UK needs all the new
supply projects to be delivering and relatively mild weather to
avoid supply shortages. If the increased Interconnector capacity
is not available (it is only planned to come on line at the end
of December) or we experience cold weather the UK can be expected
to hit significant supply shortages during the coming winter.
SOURCE INFORMATION/REFERENCES
This analysis draws on information from the DTI and Transco
websites.
DTI table 1.1.8 is used for temperature data going back to
1970.
DTI table 4.2 is used for supply and demand data. This has
been compared to data from the Transco Information exchange website
to validate, where possible, the data to use.
The ICUK website has been used as a source of historic flow
data from the Interconnector.
Independent information has been used as a source of Norwegian
imports.
Transco Winter Outlook Report—October 2005.
| 
