Sunrise LNG in Timor-Leste: Dreams, Realities and Challenges
A Report by La’o Hamutuk
Timor-Leste Institute for Reconstruction Monitoring and Analysis
Appendix 3. Fiscal effects
As the main text explains, the effect on net revenues to Timor-Leste’s government of where the LNG plant is located will depend on the effect of that decision on the sales price of the LNG it produces and on the cost of constructing the pipeline and liquefaction plant. It is also determined by the tax schedules that would apply to upstream and downstream projects. This appendix describes the assumptions on which the estimates presented in the main text are based. The calculations themselves can be inspected (and other assumptions attempted) in the spreadsheet accompanying this report, available at http://www.laohamutuk.org/Oil/LNG/FiscalBenefits.xls.
The upstream revenue to Timor-Leste’s government can be divided into three components. According to the IUA, Sunrise is shared by Australia and JPDA in the proportion 79.9%-20.1%. Timor-Leste’s first source of revenue from Sunrise is its 90% share of JPDA’s take of profits in its part of Sunrise as given by the Sunrise Production Sharing Contract (PSC). Secondly, Timor-Leste imposes income taxes on its share of the JPDA share (90% of 20.1%) of the upstream companies’ income after they have shared production with JPDA. Finally, the CMATS Treaty provides that Australia should make transfers to Timor-Leste in the amount necessary to equalize the two governments’ revenues from the upstream project.
The downstream revenue to Timor-Leste can be divided into two components. The first consists of income taxes on the LNG plant itself. The second consists of various withholding or wage taxes on money that is spent by the LNG plant—on salaries or locally sourced supplies—but these taxes are not paid by the plant, but rather by those who receive them (we call these “Other Timor-Leste taxes” in the tables). We have not attempted to estimate any multiplier effects in the local economy—we only estimate the number of jobs and the magnitude of purchases that the LNG plant itself may be expected to carry out. We have also left out any revenues from import duties or other indirect taxes, as we cannot confidently estimate what they might be, and they would in any case be too small to affect our overall analysis.
We have attempted to model upstream and downstream aggregate cash flows, upstream revenues for Timor-Leste, Australia and the companies, and downstream revenues for Timor-Leste and the companies if the LNG plant comes to Timor-Leste. We have also tried to compare the two land-based options for the LNG plant and the pipeline (Timor-Leste and Australia), but it is beyond the scope of this report to model the domestic Australian taxes (including taxes paid by people and businesses employed by or supplying the LNG plant) if the LNG facility is in Australia. In the comparisons, we therefore restrict ourselves to reporting total downstream project revenues (company profits plus taxes paid by the project) for the Australian option, and have made the simplifying assumption that Australia’s tax revenues from the downstream project are just the flat 30% company tax applied to the total profits.
We would like to emphasize that the numbers used here are often quite tentative. Estimates of costs and revenues in a large petroleum project are necessarily marred by uncertainty, and we have not had access to more than easily available public information regarding the prospects of Sunrise. Further, the overlapping legal and taxation regimes that results from the Indonesian occupation and the UN Transitional Administration, as well as the complicated situation in the Timor Sea, make it very difficult to be certain of the exact tax and profit-sharing rules that would apply to the project. We therefore caution readers to double-check our numbers before making use of our results.
All monetary values are denominated in United States dollars. Totals over time are given in 2006 net present values, discounting only by the inflation rate, not by any further discount rate.
Different petroleum products are quantified in different units. Natural gas is usually measured by volume (in cubic feet or cubic meters) but priced by energy units (measured in British thermal units); LNG is measured in metric tons; and crude oil and other condensates in barrels. Different types of petroleum can be compared by their energy content, which is measured in barrels of oil equivalent.
Table 13. Approximate conversion factors
Natural gas (NG) and liquefied natural gas (LNG)
billion cubic meters NG
billion cubic feet NG
million tons oil equivalent
trillion British thermal units
million barrels oil equivalent
——————— Multiply by —————————
1 billion cubic meters NG
1 billion cubic feet NG
1 million tons oil equivalent
1 million tons LNG
1 trillion British thermal units (BTU)
1 million barrels oil equivalent
Source: BP Statistical Review of World Energy, June 2007 
Our different scenarios are explained in the Fiscal Effects section of Chapter 4. In the following tables, we give details of how we have chosen the different parameters. As explained in Box 12, costs in liquefaction plant and pipeline construction have recently escalated strongly. In our scenarios, we use what we believe were reasonable price estimates before these increases. We still think these costs estimates are plausible in the long term, given that the current high prices are driven by particularly tight market conditions, rather than physical or technological constraints. For completeness, however, we also provide calculations for cost parameters three times as large as what we use in the baseline moderate scenarios.
Furthermore, our scenarios assume that any downstream project in Timor-Leste would be subject to the domestic taxation system at the time of writing. As explained in Box 11, however, a reform proposal for domestic taxes is currently being considered by the Timor-Leste government. We therefore also include in the spreadsheet additional calculations using the new proposed tax rates, one example of which is in Scenario 1a. Note that the tax reform does not make a difference to the upstream project.
Table 14. Main assumptions (in “Assumptions and Results” worksheet)
Explanation of assumption
Public estimates have varied from 7.68 trillion cubic feet (tcf) to more than 9 tcf. Privately, Timor-Leste government officials have expressed their belief that Sunrise may contain more than 11 tcf. We have conservatively used an estimate of 8.35 tcf from the middle of the range of publicized figures. The real figure is likely to be higher if gas prices remain at high levels, since that would make it economical to use more expensive technology to recover more gas. There are also 300 million barrels of condensate (see Box 5). We do not calculate the economic effect of condensate production, as condensate can be exported directly from the offshore facilities, regardless of where the LNG plant is located. So condensate production does not affect the comparison between different solutions for the liquefaction process.
Upstream exploration costs
Woodside reports having spent about US$185 million (Australian $250 million) on prospecting and other preparation for Sunrise so far. We assume that the total exploration and other preparatory costs may come to US$300 million in total, slightly higher than McKee’s estimate.  While this figure is a very rough estimate, it has little effect on the calculations.
Upstream construction costs
ACIL Consulting  assumes total upstream investments of US$1,150 million (Australian $1,500 million) for a 5 mtpa-capacity extraction facility and US$1,500 million (Australian $2,000 million) for a 6.6 mtpa-capacity facility. Our assumptions for LNG production capacity and feed gas lost to electricity consumption in the liquefaction process (see below) suggest the higher end of the range. Given our assumptions for exploration (see above) and decommissioning costs (see below) for the upstream project, we calculate that the remaining offshore facilities will cost US$1 billion. For simplicity, we assume that construction costs are borne through the first four years of the project (so we ignore construction expansions after production has commenced). This simplification does not significantly affect the results. It does, however, mean that our costs are classified somewhat differently from what may be found in other industry calculations. “Upstream construction costs” includes costs that have to be expended before production can start. “Upstream operating costs,” below, includes what economists refer to as variable costs, meaning all those costs that depend on production volume. This last item may include construction, for example when a new well is drilled after production has started. This partitioning of cost may make our “construction costs” seem lower than other estimates of “capital expenditures,” whereas our “operating costs” may seem similarly higher than other estimates of “operating expenditures.” We advise readers who compare our costs with other sources to keep this in mind, and to look at total costs to get a simple comparison of which estimates are more conservative.
Note also that this figure for “upstream construction costs” does not include the cost of laying the pipeline, which we attribute to the downstream project.
Upstream operating costs
We assume the non-fixed capital costs of gas extraction to be US$0.50/MM BTU, equivalent to US$2.94/BOE. This figure corresponds to the lower end of the estimate range given for exploration and production costs in Foss. 
Upstream decommissioning costs
The U.S. Government provides estimates for decommissioning of petroleum production rigs off the U.S. Pacific Coast.  Their numbers suggest that the facilities of the size Woodside  envisages at Sunrise (39,400 tons) would cost about US$85 million (in 2004 dollars) to dispose of. Accordingly, we assume upstream decommissioning costs of $100 million. This can be compared with McKee’s estimate of $150 for the whole production chain (upstream and downstream). 
Total fixed costs
The sum of exploration, construction, and decommissioning costs.
Total per unit produced
Total fixed costs divided by the total quantity of extractable gas, plus the operating costs per unit of gas.
For simplicity, we assume the same timeline for the construction of both the upstream and downstream facilities, four years. For comparison, the Darwin LNG plant (www.darwinlng.com) was built in less than three years (although the upstream started several years earlier), while Snøhvit LNG took more than five (www.snohvit.com). Atlantic LNG’s first train took about three years to construct, and some of the expansion trains even less.  It seems reasonable to expect a longer construction time in Timor-Leste since some infrastructure will need to be built before plant construction can begin, and the various components have to be constructed off-site and shipped in. This could take a long time, especially with the long waiting times in today’s petroleum-related construction sector, which faces very high demand for exploration and other equipment.
Number of years of operation
This is calculated from the amount of feed gas needed for the production of the assumed quantity of LNG and to power the electricity generation plant. We simplify by assuming that full capacity is reached in the first year, and is constant so long as there is gas remaining in the reserves. Including a ramp-up and ramp-down period would not significantly change the results we examine in the report.
Yearly LNG output
No decision seems to have made about capacity. Numbers that have been mentioned range from 5 mtpa to 7 mtpa. In our calculations we assume 5.3 mtpa of LNG produced, the quantity proposed by Sunrise operator Woodside in official presentations.  Given the assumption that 9% of the feed gas is used to power the liquefaction process (see below), this means a total of 287 bcf of natural gas must be pumped to the LNG plant annually, or 0.78 bcf/day.
Percentage of gas liquefied
A liquefaction facility would need to be fully powered by its own power plant to supply the energy needed to pump and liquefy the gas. This plant would consume a share of the gas purchased from the upstream project. The gas consumed in the liquefaction process at Darwin LNG is 7%, and Woodside has projected a 9% figure for Sunrise.  We use the 9% assumption in our calculations.
Construction costs, pipeline
The feasibility and costs of laying a pipeline from Sunrise to Timor-Leste is a matter of great dispute. While the distance from Sunrise to Timor-Leste is much shorter than to Australia, a pipeline to Timor-Leste would have to cross the Timor Trough with depths down to 3,000 meters. Woodside has dismissed that option and wants the gas from an upstream Sunrise project to be pumped to Australia for processing. Several observers, however, think a pipeline to Timor-Leste could be significantly cheaper than the alternative of a pipeline to Darwin. ,  A study by Intec estimates the cost of a pipeline of sufficient size for a 6 mtpa LNG plant from Bayu-Undan to Timor-Leste’s southern coast—a similar distance as that from Sunrise to the coast—to be US$317 million.  (They also estimate that a pipeline could be laid to connect Sunrise to Bayu-Undan for $171 million.) On the other hand, an unpublished study by Woodside from 2004 estimates the cost for a Sunrise-Timor-Leste pipeline at $721 million, compared to a pipeline to Darwin for $566 million. Another consultant’s study, carried out for the government of Timor-Leste but also not published, criticizes Woodside’s numbers and estimates a pipeline to Timor-Leste could be laid for as little as $448 million.  This report also suggests that Woodside’s estimate for the Darwin option understates the true costs.
We are not in a position to make a definitive judgment about what a pipeline would cost. In our comparisons, we use two possible cost levels: $550 million for the “moderate” option and $750 million for the “expensive” option, in line with the lower and upper ends of the range of estimates that have been discussed. In practice, the cost will also depend in part on the capacity of the plant, since that influences the required diameter of the pipeline, but this makes a negligible difference to our results, so we simplify by assuming a fixed cost. We compare the two pipeline options for three assumptions: that they are equally costly, that the Timor-Leste option is the cheaper one, or that the Australia option is the cheaper one.
Construction costs, liquefaction plant and associated facilities
Recent literature suggests that a cost of US$250 per tpa capacity is achievable, and even lower costs for expansions of existing projects. As an illustration, train 1 of Atlantic LNG in Trinidad and Tobago was built at less than $200/tpa in 1999 ; the Train Four expansion was reported in 2005 to amount to $1.3 billion for a capacity of 5.2 mtpa.  Darwin LNG cost just over US$1bn (Australian $1.4bn) for a nominal capacity of 3.5 mtpa, according to the contractor Bechtel, which amounts to just under $300/tpa (http://www.bechtel.com/darwin_lng.html).
It is disputed which of the two onshore options for the LNG plant would be cheaper to construct. A plant in Timor-Leste would be a “greenfield” development, and would perhaps face a higher cost of capital because of a higher perceived risk. In Australia, on the other hand, it might be possible to expand Darwin LNG as a “brownfield” development, but some commentators argue that Australian labor regulations which increase labor cost and hinder the use of modular construction techniques.  Even an expanded Darwin LNG plant might not be available to process Sunrise gas, as other nearby gas fields might have a prior claim on the facility, which cannot be expanded beyond 10 mtpa because of site constraints.
For our scenarios, we use $250/tpa for the “moderate” assumptions and $300/tpa for the “expensive” assumption. The “very high cost” assumption in Scenario 2a is based on $750/tpa.
The assumption that the cost is proportional to capacity is a simplification; in practice, there will be a fixed component to the construction costs due to minimum infrastructure work that has to be done regardless of the capacity. This simplification does not have a significant effect on our results.
Total downstream construction costs
The cost of the pipeline plus the cost of the LNG plant.
Respective share of materials and labor in construction costs
According to a recent survey, approximately “half of [the cost of a liquefaction plant] is for construction and related costs, 30% is for equipment, and 20% is for bulk materials.”  A rule-of-thumb is that roughly 50% of the costs are spent on materials and 50% on actual contractor work, which under current Timor-Leste domestic tax laws is taxed differently from other income. We assume the share of materials to be relatively higher for the pipeline compared to the liquefaction plant, and therefore assume a 60% share of the overall construction costs for materials and the remainder for labor (“construction services.”)
Dollar amount obtained by multiplying the total construction costs by the share of labor in costs. (This spending is subject to a special tax under Timor-Leste’s domestic tax regime.)
Percentage to Timorese residents
This is a very difficult quantity to estimate, but will by all reasonable assumptions be small, given the poor state of development of industry in Timor-Leste. According to Atlantic LNG’s numbers, about 18% of the total construction costs for the Train Four expansion went to Trinidadian firms. In Timor-Leste, virtually none of the materials would be purchased locally, and a significantly lower portion of construction services would be locally sourced than in Trinidad. We assume that 10% of the expenditures on construction services (and none of the materials costs) are spent on Timorese residents. This figure, however, will depend greatly on the considerations examined in Chapter 5.
Operating costs per million BTU (MMBTU)
Since the calculations assume that power generation is fueled by feed gas from the Sunrise reserves, we assume a low remaining unit cost of US$.30/MMBTU for the operating expenses of LNG liquefaction (equivalent to about $15 per metric ton of LNG). This brings the total unit cost for the downstream project (including construction costs but excluding power) to about $.80/MMBTU or $40/ton, which is at the lower end of generic estimates (e.g. Foss  estimates total liquefaction costs to be in the $.80-$1.20/MMBTU range; EIA  cites an estimate of $1.09/MMBTU for a greenfield project of 8 mtpa capacity).
Share salaries to residents, to non-residents, and to local content supplies.
These are also difficult quantities to estimate, but given the small number of personnel needed for plant operation (see Chapter 5), it is likely that most of the operating costs will involve non-labor costs. While most of the permanent personnel will be Timorese residents under the extant domestic tax laws, some salary costs would be paid to short-term, but high-paid technical experts. We assume the resident salary share of the operating costs to be 5% and the non-resident salary share to be 10%. Of non-labor costs, we also assume that only a small share will be purchased locally, at 10% of the total. Of course all of these figures will depend on the considerations discussed in Chapter 4 and Chapter 5.
This is a very uncertain assumption, as costs of disposal would be very sensitive to the kind of environment that would have been affected, and the state to which it would have to be restored. Pipeline removal is very expensive: U.S. Government  estimates a cost of more than $100,000 per mile of pipeline for disposal, but also around the same amount per day of on-site operations for retrieval of the pipeline, which can take 5-7 hours for a 120-foot (36-meter) cut. This would amount to some $1.5 million per kilometer of pipeline. We cannot estimate how much of the pipeline would have to be removed, but enough should be removed to remove any hazard to the environment, in particular near the coast. In addition, the processing plant itself would have to be dismantled in a safe way, although many of the port facilities could probably be transferred to other uses. Our calculations assume decommissioning costs for both pipeline and onshore facilities in the amount of US$200 million, as we do not believe Timor-Leste ought to risk damage to its environment for what would be a relatively small increase in company profit and tax revenues. Nevertheless, we emphasize that these are very uncertain numbers.
Total fixed costs
The sum of construction and decommissioning costs.
Total costs per unit of gas sold
Total fixed costs divided by the total quantity of LNG sold over the lifetime of the project, plus the downstream operating costs per unit of LNG. This reflects economic unit cost of the downstream operations (not including the upstream costs).
Allowed internal rate of return
The IUA sets out principles for determining the price at which petroleum is sold by the upstream to the downstream project (the netback price). The price has to be based on arm’s-length principles, which means a price that would be agreed if the upstream and downstream projects were run by different companies not trying to collude. What this price might be is a matter of speculation, but the IUA provides a rule for the downstream project’s internal rate of return (IRR) in the absence of arm’s-length transactions (for example if some of companies from the Sunrise consortium are also involved in the downstream project). The posited IRR is set to 10.5% for a land-based LNG plant. We use this to calculate how the netback price will be set.
Note that the IUA formulation is vague on whether the 10.5% is a real or a nominal rate of return. An advisor to the Timor-Leste government informs us that it should be a nominal rate of return, which would make the project much less attractive to investors if the world returned to a high-inflation (or even moderate inflation) environment. In such a situation, the netback price would likely be renegotiated, and the IUA explicitly provides for this possibility. We therefore simply assume a constant low inflation rate of 2.6% (the difference between inflation-adjusted and non-adjusted U.S. Treasury bonds at the time of writing), and expect that any significant change in the inflation rate would cause the nominal IRR to be renegotiated to maintain the same real (post-inflation) IRR.
LNG sales price
For convenience, the spreadsheet also reports the corresponding price per barrel of oil equivalent, using energy content to calculate the equivalent price. In our high-price scenarios, we use US$7.50/MMBTU, equivalent to about $45/barrel of oil equivalent. This is somewhat lower than the longest-term future price contracts traded on the NYMEX commodity exchange, which captures the commodity’s markets average guess at the future price. The average price of all future natural gas contracts traded on NYMEX—monthly contracts from December 2007 through December 2012—was $8.35 on 2 November 2007. The average price for the 12 contracts for 2008 delivery was $8.54, and the average price for the 12 contracts for 2012 delivery was $7.96. At these price levels, our assumed LNG sales price therefore allows for shipping and regasification costs.
We note that these prices are historically high (see Figure 5). The average spot price for natural gas in the U.S. was US$5.76 during 2001-2005 and $3.28 in the 1996-2000 period (in 2005 dollars, calculated from the IMF’s International Financial Statistics). It is notoriously difficult to predict energy prices, but we cannot rule out that in the long term, they will return to the very low levels of the late 1990s. We therefore also include a set of low-price scenarios, where we assume a natural gas price of $3.50/MMBTU, equivalent to about $21/barrel of oil equivalent.
We assume a 2.6% inflation rate, the difference between yields on U.S. Treasury 30-year bonds with and without inflation indexation at the time of writing. (http://www.bloomberg.com/markets/rates/index.html quotes yields of 1.98% versus 4.62% on 2 November 2007.)
Long-term borrowing rate for augmentation
This is the base rate for the “augmentation rates” that Australian petroleum taxation adds to capital and operating costs. We assume a real interest rate of 5%, so that the long-term borrowing rate is 5% plus the rate of inflation. For Australian petroleum taxation, see http://www.ato.gov.au/large/pathway.asp?pc=001/009/029&cy=1.
Table 15. Upstream assumptions (in “Upstream” worksheet)
Explanation of assumption
Quantity of gas purchased by the downstream project
Given by the assumptions of yearly output and liquefaction energy consumption.
Price of gas purchased by the downstream project
|The netback price, calculated to satisfy the required IRR.|
The rates for the upstream exploration, construction and operating costs are explained in Table 14 “Main Assumptions.” We assume very simple time profiles for exploration and construction, with equal exploration costs over two years, and a ramped-up four-year construction period. While not quite realistic, this makes no difference to our analysis.
Upstream project net cash flow and IRR
The difference between gas sales revenues and upstream costs, and the resulting IRR from that cash flow (a measure of profitability).
JPDA production sharing rules
As per production-sharing contract terms preserved under Annex F of the Timor Sea Treaty. The PSC for Sunrise has not been made public, but the relevant terms have been communicated to us.
Share of capital costs in construction
We assume that a large part of construction costs will be for equipment and materials rather than labor, and set this to 75%.
Australian take of JPDA share
From Australian tax rules. Note that the Petroleum Resource Rent Tax (PRRT) does not apply to Australia’s share of JPDA revenues. 
Timor-Leste take of JPDA share
Preserved Indonesian income tax rules. We approximate the depreciation rules with 20-year straight-line depreciation.
Australia’s (non-JPDA) share of upstream cash flow
Australian tax rules, including 40% Petroleum Resource Rent Tax.
CMATS provides for a transfer from Australia to Timor-Leste of the amount that would equalize the revenues the two governments receive from the upstream Sunrise project.
Table 16. Downstream assumptions (in “Downstream” worksheet)
Explanation of assumption
Inflation and real cost factors
The inflation factor is calculated by compounding the assumed constant rate of inflation inputted into the Main Assumptions sheet. In our scenarios, we assume that the natural gas price is constant over time in real terms (that is, it just increases by the rate of inflation in nominal terms). For the convenience of users, however, we include a line where users may experiment with assumptions of increasing or decreasing real natural gas prices. The real rate of increase should be inputted in the first (left-most) cell of line 11 (to do this, the user must “unprotect” the sheet in the Excel “Tools” menu).
Construction and operating costs
See the assumptions for the costs in Table 14 “Main Assumptions.” For the time profile of cost, we posit a simple ramped-up time profile with peak construction activity in the third year.
Sales of LNG and purchases of feed gas
See the assumptions in Table 14 “Main Assumptions.” The netback price for purchases of feed gas (line 19) is calculated in lines 26-52. Line 23 adjusts the assumed natural gas price by the inflation factor (from line 10) and the real natural gas price change factor (line 11). We assume the latter to be constant in our scenarios.
Calculation of netback price
Based on formula in IUA Annex 6. Lines 27-31 give the decomposition of costs as is Annex 6. Lines 34-38 give the same costs discounted at the required IRR (given in Main Assumptions). Line 40 repeats the cost of gas purchased, while line 42 gives the net cash flow for the downstream project (line 23 minus line 31 minus line 40). Line 43 gives the same number in constant (year 0) dollars, i.e. adjusted for inflation.
The actual IRRs for that cash flow (nominal and real) are given in lines 45-46, as a check that the calculated netback price produces the right IRR. The netback price is calculated by rearranging the formula in Annex 6 as follows:
where VDP is the total market value of the LNG (line 27), ECC expenditures for capital costs (line 28), OC operating costs (line 29), CDC costs for decommissioning (line 30), and QH the quantity of gas purchased (line 18). The resulting netback price is given in line 50. Note that this is a fixed nominal price, whose real value will change with inflation. For comparison with the LNG sales price, it is more informative to look at the real (year 0) price, which is given by the net present value of the money paid to the upstream by the downstream project for the feed gas (line 20) divided by the total quantity of gas bought (line 18). This figure is given in lines 51 and 52.
Taxes to Timor-Leste
As explained in Chapter 4.2, we calculate downstream tax revenue to Timor-Leste according to the current regime of business taxation. The estimates assume that the regime will not change, and in particular that no tax incentives are given in order to attract the project to Timor-Leste. We assume that construction costs are evenly distributed into the two depreciation categories of “buildings” and “machinery.” When contractors have the choice between depreciation methods, we assume they choose the one that maximizes their IRR. The assumptions on the share of construction and operating costs going to local content are given in Table Main Assumptions. Finally, we assume that 50% of local supply purchases indirectly accrue to wages to residents. Note that we are ignoring indirect taxes (customs duties and excise taxes) as we cannot confidently predict how these will apply to the LNG plant.
In a separate sheet, “Downstream (new tax regime),” we model the downstream tax revenues to Timor-Leste that would result if the tax reforms under consideration by the government at the time of writing (see Box 11) are enacted as they are currently proposed. We have included this for completeness, but underline that we have not been able to thoroughly research these rates, and based these estimates simply on short publications by the government.  The calculations are in lines 56-90, and also ignore indirect taxes.
Approximation of downstream taxes to Australia
We use the model of Australian taxation of Australia’s share of JPDA revenues as a simple approximation of company taxation and depreciation rules that might apply to an Australian LNG plant. These figures should be seen as very approximate, as it is beyond the scope of this report to do an in-depth analysis of Australia’s domestic taxation.
The sheet “Downstream (new tax regime)” reproduces the same estimates in lines 95-108.