Pramod Kumar, Accountant Member. -

1. By way of this appeal, the tax-payer has called into question correctness of CIT(A)’s order dated 30-3-2004, for the assessment year 1997-98, on the following grounds :

“1. The learned CIT(A) erred in upholding the action of the Assessing Officer in charging to tax receipts from all the contracts on the ground that the appellant had a permanent establishment in India, ignoring the fact that the duration of work performed under each of the contract did not exceed nine months as per India Mauritius tax treaty.

2. The learned CIT(A) ought to have held that no income could be attributed to the permanent establishment.

3. The learned CIT(A) ought to have held that no income could be attributed to the permanent establishment may be taxed in India.”

2. The relevant material facts, as culled out from the material on record, are as follows : The assessee before us is a Mauritius based company and is engaged in the business of installation of offshore platforms, decks, pipelines, jackets and various other similar activities, for the purpose of mineral oil exploration. During the relevant financial period, the assessee was engaged in executing certain installation contracts in Tapti and Panna offshore fields in the Indian continental shelf. However, according to the assessee, the income so earned by the assessee was not liable to be taxed in India. The assessee claimed the treaty benefits under India Mauritius Convention for Avoidance of Double Taxation - [1984] 146 ITR (St.) 214. According to the assessee, its income from the contracts so executed was in the nature of business profits which could be taxed only under article 7 of the Convention, but existence of a permanent establishment, as defined under article 5 of the said convention, was a sine qua non for such a taxability. It was further pointed out that since the assessee did not any such permanent establishment in India, the income of the assessee could not be brought to tax in India. In support of this claim, it was pointed out that the appellant had executed following contracts during the relevant period :

3. It was also pointed out by the assessee that, in terms of the provisions of India Mauritius Tax Convention, a building site or construction or assembly project, or supervisory activities in connection therewith, shall be regarded as a permanent establishment only where such site, project or supervisory activities, in connection therewith, continue for a period of more than nine months. This time-limit, according to the assessee, is to be applied to each of the activity separately. It was also submitted that various periods during which the work is carried on in India cannot be aggregated to apply this test of nine months’ period. None of these submissions impressed the Assessing Officer. He rejected the claim of the assessee by observing as follows :

“I have considered assessee’s explanation, in the facts and in the circumstances of the case, the issue decided as under :-

On perusal of the contract No. 95/C/02 and contract No. 95/C/04, referred by the assessee’s reply as contract No. D-4507 and D-4522 seen that the first phase of Panna field was executed from March 1996 to 22nd November, 1996 and the second phase had commenced from 1st November, 1996, completed on 18-5-1997. The duration of the 2 contracts as per para-2 of the contracts are described below :—

Contract No. 95/C/02 (D-4507)

Duration of contract :

The term and duration of this Agreement is based upon the following stipulations. The duration of the Agreement will be as follows :

The contract duration shall be two months from the commencement date plus any time extension approved by Enron.

Commencement date shall be on or about March 16, 1996.

Completion date is as specified in Exh. “H”.

Contract No. 95/C/04 (D-4522)

Duration of contract :

The term and duration of this agreement is based upon the following stipulations :

Commencement date November 1, 1996.

Completion date is as specified in May 31, 1997.

In view of the duration clause of the contracts and completion certificates issued by the EOGI Ltd. It is apparently clear that the work in Panna field was executed from March 1996 to May 1997. The assessee has deliberately given wrong dates of during of work in both the return of income and in reply dated 15-3-2000. Invoices have also been collected under section 133(6). The last invoice for Panna field is raised on 4-5-1992 (sic 4-5-1998) (Invoice No. 4522607/LO) and first invoice was raised on 1-11-1996 (Invoice No. 4522001/LO) under contract No. D-4522. The invoices are raised in 19 months under this contract.

In view of the above facts, it is clear that the assessee work in Panna field has exceeded the period for more than 9 months and the assessee is Permanent Establishment in India. The completion certificate issued by assessee on 18-5-1997 is only in respect of a part of contract No. D-4522.

It is pertinent to mention that the assessee-company has been engaged in offshore exploration activities in Indian waters since financial year 1994-95. The assessee had claimed exemption from Permanent Establishment, only on the basis of claim that the duration of the execution of the various projects was less than 9 months. While computing the period of presence, the assessee has not taken into account the period spent on pre-job inspection of the site mobilization of personal and vessels before execution of contract and demobilization of the personnel and vessels after execution and winding up activities. These all activities forms the integral part of the contract performed by the assessee in Indian waters, hence, it is to be taken into account for working out the duration of the project.

The issue of duration of oil drilling project has arisen before French Counseil d Etat in connection with the treaties with Spain and Algeria (33 Droit Fiscal 1, 111, wherein, the Spanish Company has drilled many holes each enduring for less than 12 months. It was held that since the aggregate period of various projects exceeded 12 months, it will constitute permanent establishment.

In view of the above discussion, it was held that the assessee-company has permanent establishment in India.”

4. Aggrieved by the stand so taken by the Assessing Officer, assessee carried the matter in appeal before the CIT(A). In addition to the legal submissions, following facts were also set out before the CIT(A) :

1.Your honour’s specific attention is drawn to the following facts :

uThe ENRON contract (D-4507) was in respect of Transportation and installation of platforms and facilities for the TAPTI AND PANNA FIELDS, Offshore India. It began only on 12 March, 1996. The ENRON contract (No. D-4507) mentions that the work shall commence on or about March 18, 1996. The completion date as per the contract is 16th May, 1996 and ended on 22 November, 1996 which is less than 9 months.

uThe work on the HEERMAC contract began on 12 November, 1996 and ended on 21 November, 1996.

uThe contract with ENRON (D-4522) was in respect of Transportation and installation of platform and facilities in the TAPTI FIELD, Offshore India. The contract mentions the commencement date as 1st November, 1996 and the completion date as 31st May, 1997. However, in actual fact the work at the site commenced only on 5th February, 1997 and was completed on 19th May, 1997.

uAt all completion dates are based on the contract completion certificate issued by the contractor and were produced to the JDIT.

uNo work was carried on in the PANNA FIELD (contract D-4522) during the period December, 1996 to January, 1997. Work on contract D-4522 did not commence until 5th February, 1997. Therefore, it cannot be said that the work was carried out on the PANNA FIELD continuously from March, 1996 to May, 1997 as claimed by the JCIT.

uThe appellant has retrieved the activity reports from its archives. Copies of relevant reports are attached to support the appellant’s submission on the physical presence of equipment and personnel of India.

2.It is pointed out that each of the contracts entered into by our clients with various parties are independent contracts and are not related to each other. Each of the contracts continued for specific duration only. It is further contended that each of these contracts, D-4406 and D-4522 was bid negotiated and awarded separately. The work scope of each contract is distinct from the others and there is no interdependence between the contracts which would lead to the conclusion that they were no more than different phases of the same project.

3.However, the JCIT has completely ignored the contract completion certificates issued by the third parties. He has proceeded only on the basis of the dates appearing in the agreement and has held the work in the Panna Field was executed from March, 1996 to May, 1997. He has erroneously held that the certificate for the ENRON (D-4522) contract is only in respect of part of a contract rather than being specific to the completion of an independent contract. He has held that the contract with ENRON for carrying out work in the TAPTI and PANNA FIELDS and on the TAPTI FIELD were merely Phase-I and Phase-II of the same contract. He has failed to appreciate that these were separate contracts and the work was carried out at various locations. He has also failed to appreciate that the contract completion certificates were in respect of the entire work carried out in India in relation to each independent contract and not for part of the work performed on a single contract.

4.He has also misquoted the dates of the invoices. He has mentioned that the last invoice for PANNA FIELD WAS RAISED ON 4-5-1992 (sic 4-5-1998) AND THE FIRST INVOICE WAS RAISED ON 1-11-1996. In fact, the invoices for contract D-4507 were raised on 3-4-1996 and 30-11-1997 respectively. The first and the last invoices for contract D-4522 were raised on 1-11-1996 and 4-5-1998 respectively. Please note that the invoice No. 4522001 dated 1-11-1996 was towards a mobilization advance and not for any activity in India. Furthermore, the next invoice No. 4522002 was dated 5-3-1997 for sail out of the Derrick Barge from outside Indian waters.

The JCIT has incorrectly used these invoice dates as his basis for determining whether a permanent establishment existed. These invoice dates are, of course, irrelevant for this purpose. Whether the appellant has a permanent establishment in India for each contract must be determined based on physical presence in India not invoice dates.

5.The JCIT has also held that while computing the presence your appellant has not taken into account the period spent on pre-job inspection of the site, mobilization of personnel and vessel before the execution of the contract and de-mobilization of the personnel and the vessels after the execution and winding up activities without any evidence on record. He has held so merely on suspicion, surmise and conjecture.

6.It is submitted that under the Indo-Mauritius tax treaty, a building site or construction or assembly project or supervisory activities in connection therewith shall be regarded as a permanent establishment where such site, project or supervisory activities continue for a period of more than nine months. In computing the period of nine months each site, project or supervisory activity is to be taken separately and the test of nine months has to be applied to each particular activity separately. The various periods during which the work was carried in India under various contracts at various periods during which the work was carried in India under various contracts at various locations cannot be aggregated to determine whether your appellant carried out the work in India for a period of more than nine months.

7.Moreover, the period of 9 months has to be computed with reference to the number of days the activity was carried out in India. In computing the period of nine months, the work, if any, done outside India cannot be considered. Your appellant has in computing the period of nine months, taken into account the time spent by them in India in performance of the work. The entire work relating to the contract was completed (including period spent on pre-job) inspection of the site and winding up activities) within a period mentioned above.”

5. The CIT(A), however, was of the view that all the time spent on executing all the contracts was to be aggregated to apply the threshold limit of nine months for ascertaining whether or not the assessee had a permanent establishment in India. It was noted by her that the assessee had carried out the work for the same contractor, either directly or indirectly, and at the same location. She noted that the work carried out by the assessee for HEERMAC was ultimately for Enron Oil & Gas Ltd., as HEERMAC was only a contractor for Enron Oil & Gas Ltd. She was, thus, of the view that the contracts were of the same nature and at the same geographical location. Since the aggregate period for the contracts so executed by the assessee exceeded the period of nine months, according to the CIT(A), it was to concluded that the assessee had a permanent establishment in India. She, thus, upheld the action of the Assessing Officer, and declined to interfere in the matter. In her rather brief order, learned CIT(A) observed as follows:

“4.16I have carefully considered the submission made by the ld. counsel and the order of the Assessing Officer as well. I have also perused the contract Nos. D-4507, D-4406, and D-4522. The point for consideration is whether each contract should be taken separately or together for the determination of PE under Indo-Mauritius tax treaty. The contention of the appellant is that each contract was distinct from the others and there was no interdependence between the contracts. I don’t agree with the contention of the appellant. If we see the facts of the case it is noticed that all the three contracts have been award to the appellant- company by one entity, i.e., Enron Oil & Gas Ltd. Two contracts D-4507 and D-4522 have been awarded directly to the appellant-company and one contract D-4506 was awarded to HEERMAC a Netherlands company which again sub-contracted it to the appellant-company. In short, ultimately all the three contracts are being carried out by the appellant-company and the principal is only one entity, i.e., Enron Oil & Gas Ltd. The appellant company is engaged in the business of transportation and installation of platform and facilities for the TAPTI AND PANNA FIELDS, offshore India. So the nature of job for all the contracts and the geographical location is also same. If we see the duration of work, it is clear that the work was carried out continuously for March, 1996 to May, 1997, of course, there was negligible interruption but, there is no overlapping of months or dates.

4.17Under article 5 of the Indian Mauritius tax treaty, a building site or construction or assembly project or supervisory activities in connection therewith shall be regarded as a Permanent Establishment where such site, project or supervisory activities continue for a period of more than 9 months. In view of the above facts, the appellant’s contention that the periods during which work was carried out in India under various contracts cannot be aggregated for determination of PE because all three contracts are commercially separate is not acceptable. In this respect Para-18 of the commentary on OECD model treaty is relevant wherein it has been observed that a “building site should be regarded as a single unit, even if it is based on several contracts, provided that it forms a coherent whole commercially and geographically”. In the appellant’s case principal is same, geographically the place is same, i.e., only one location, i.e., Tapti and Panna Field Offshore, nature of job is also same. All the contracts are ultimately carried out by the appellant’s company only. If we see the duration of work in India it is noticed that the work was started in the month of March and continued till May, 1997. Thus, there is no overlapping of period also. The work has been carried out continuously from March 1996 to May 1997.

4.18In the light of the above facts, I hold that duration of work in India has exceeded 9 months and the period for all the contracts should be taken together under article 5 of the India Mauritius treaty as it constitutes PE. Since the computation of the profit and details of expenditure attributable to PE has not been furnished by the appellant, hence, the basis for taxation adopted by the Assessing Officer who has taxed the income for the appellant under section 44BB of the Income-tax Act is confirmed. Appellant’s appeal on these grounds is dismissed.”

6. Aggrieved of the stand so taken by the CIT(A), the assessee is in further appeal before us.

7. We have heard the rival contentions, perused the material on record and duly considered factual matrix of the case as also the applicable legal position.

8. We find that there is no dispute about applicability of India Mauritius tax treaty on the facts of the present case, as also about the fundamental position that the provisions of the said treaty being beneficial to the assessee, the same will override the provisions of the Indian Income-tax Act. It is also not in dispute that the profits earned by the assessee from these contracts are business profits in nature and can only be brought to tax in India in the event of Mauritian company having been held to have a Permanent Establishment (PE) in India. The question that we must, therefore, address ourselves to is whether or not, on the facts and in the circumstances of the case, the assessee can be said to have a PE in India. Article 5 (2)(i) of the Indo-Mauritius tax treaty, which is broadly the same as article 5(3)(a) of UN Model Convention - except mainly for replacement of ‘six months’ duration test by ‘nine months’ duration test, and for including it in paragraph 5(2), which is generally required to be read along with paragraph 5(1), as against paragraph 5(3) in the UN Model Convention which is more of an agreed extension of PE definition rather than being illustrative in nature. The relevant extracts from article 5 of India Mauritius tax treaty are as follows :—

“Article 5 - Permanent Establishment

1.For the purpose of this Convention, the term ‘permanent establishment’ means a fixed place of business through which the business of enterprise is wholly or partly carried on.

2.The term ‘permanent establishment’ shall include :

******

(i)a building site or construction or assembly project or supervisory activities in connection therewith, where such site, project or supervisory activity continues for a period of more than nine months.”

9. In view of the above treaty provisions, it is unambiguous that a PE refers to a fixed place of business through which business of the enterprise is wholly or partly carried on, and includes, inter alia, “a building site or construction or assembly project, or supervisory activities connected therewith, where such site, project or supervisory activity continue for a period of more than nine months.” In a way, thus, the permanence test for existence of a PE stands substituted to this limited extent, by a duration test for certain types of business activities, i.e., building construction, construction or assembly project, or supervisory activity connected therewith. There is also a valid, and more holistic view of the matter, that this duration test does not really substitute permanence test but only limits the application of general principle of permanence test inasmuch as unless the activities of the specified nature cross the threshold time-limit of nine months, even if there exists a PE under the general rule of article 5(1), it will be outside the ambit of definition of PE by the virtue of article 5(2)(i) - Hon’ble AAR’s ruling in KT Corpn. v. DIT 23 DTR 361 supports this view. Be that as it may, even a plain reading of article 5(2)(i) would show that, for the purpose of computing the threshold time-limit, what is to be taken into account is activities of a foreign enterprise on a particular site or a particular project, or supervisory activity connected therewith, and not on all the activities in a tax jurisdiction as whole. It is important to bear in mind the fact that the expressions used in the relevant definition clause are in singular, and there is no specific mention about aggregating the number of days spent on various sites, projects or activities. In other words, each of the building site, construction project, assembly project or supervisory activities in connection therewith is to be viewed on standalone basis. Broadly, the underlying rationale of this approach is that various business activities performed by one and same enterprise, none of which constitutes a PE, cannot lead to a PE, if combined. In our humble understanding, the very conceptual foundation of this approach rests on the assumption that various business activities of the enterprise in different locations are not so inextricably interconnected that these are essentially required to be viewed as a coherent whole. In a typical building site, assembly or installation project, or supervisory activities in connection therewith, each of site or project is an independent unit, and the approach to these types of PEs recognize this normal business practice. The unambiguous principle, underlying this approach, seems to be to view these business activities at different locations on standalone basis. It is also interesting to note that in certain treaties entered into by India, there is a specific departure from this rule as evident from the wordings used in definition clauses of corresponding PEs. Take for example, article 5(2)(k) of India Australia tax treaty - 194 ITR (St.) 241, which states that “The term ‘permanent establishment’ shall exclude especially . . . . a building site or construction, installation or assembly project, or supervisory activities in connection with such a site or project, where that site or project exists or those activities are carried on (whether separately or together with other sites, projects or activities) for more than six months.” [Emphasis supplied]

In the case of India Thailand tax treaty 161 ITR (St.) 82, the definition for this type of permanent establishment, which finds place in article 5(2)(h) of the said treaty, is worded as “ a building site or construction or assembly project, or supervisory activities in connection therewith, where such site, project or activity continues for the same or a connected project for a period or periods aggregating to more than 183 days.” [Emphasis supplied]

Similar are the provisions in India’s tax treaties with Austria 251 ITR (St.) 97, Belgium 247 ITR (St.) 39, Bulgaria 220 ITR (St.) 30, Canada 229 ITR (St.) 44, China 214 ITR (St.) 160, Denmark 180 ITR (St.) 1, Italy 220 ITR (St.) 3, New Zealand 166 ITR (St.) 90, Norway 169 ITR (St.) 15, Spain 214 ITR (St.) 197, Turkey 224 ITR (St.) 145 and USA 187 ITR (St.) 102. In all these cases, the relevant PE clauses are so worded that there is a specific mention for application of aggregation principle on all, or even connected, sites, projects or activities for computation of threshold duration test. Even such an aggregation, when applicable, would require exclusion of double counting of days when more than one site or project exists on a day, or when work is carried out at two or more different places on a day, as multiple counting of common days would lead to an absurdity inasmuch as when work is carried on five sites together for one hundred days each, such a computation will lead to five hundred days in a year which is an impossibility. Therefore, when definition clause specifically provides for aggregation of time spent on various sites, projects or activities, the sum total of the time spent on such sites, projects or activities, except for parallel counting of days, is to be taken into account for applying the threshold time-limit. However, when aggregation is not specifically provided for in the relevant PE definition clause, as in the present case, normally it cannot be open to us to infer the application of aggregation principle.

10. There is unanimity in OECD and UN Model Convention Commentaries at the duration test “applies to each individual site or project”. In paragraph 18 of the OECD Model Convention Commentary-OECD Model Tax Convention on Income and Capital, 1992 (as updated in 2005) to article 5, it is specifically stated so. UN Model Convention Commentary-UN Model Tax Convention between Developed and Developing Countries, 2001, in paragraph 11 of its commentary on article 5, reproduces the paragraph 18 of the OECD Commentary and adopts it. However, this replacement of or modification of - whichever way one views it, permanence test for existence of PE, by test of minimum length of time - as in the situation before us, has left scope of abuse of these provisions such as by artificially splitting the contracts, each covering a period of less than threshold limit and each attributed to different company owned by the same group. Recognizing this fact, the OECD Commentary dealing with the twelve month test for construction assembly and project site prescribed in the OECD Model Convention, observes that, “apart from the fact that such abuses may, depending on the circumstances, fall under the application of legislative or judicial anti-avoidance rules, countries concerned with this issue may adopt solutions in the framework of bilateral negotiations”. The OECD Commentary further recognizes that a building site should be regarded as a single unit, even if it is based on several contracts, provided it forms a coherent whole commercially and geographically, and that in a situation in which the very nature of construction or installation project may be such that the contractors activity is to be relocated continuously or at least from time-to-time (e.g., construction of roads and canals, dredging of waterways or laying of pipelines), the activities performed at each particular spot in a single project must be regarded as a single unit. In other words, OECD Commentary refers to the situations, in the second category, in which aggregation principle is to be applied even in the absence of specific treaty provisions to that effect. We are in considered agreement with this analysis and approach of the OECD Commentary which has also been, as we have noted earlier, adopted by the UN Commentary as well. However, there are two important issues that we need to deal with at this stage - first, as to who has the onus to show that the contracts are artificially split, or otherwise the affairs are so arranged, so as to circumvent the duration test; and - second, what are the circumstances in which the aggregation principle is to be applied, even in the absence of specific provisions to that effect in a tax treaty, so as to give a reasonable meaning to definition of a PE in respect of building, construction, or assembly project or supervisory activity in connection therewith. In our considered views, these are only two sets of circumstances in which time on each set of relevant business activity by an enterprise, in the other Contracting State, is to be aggregated.

11. As for the cases of alleged treaty abuse, alleged artificial splitting of contracts, or other alleged modes of maneuverings to enter into sham arrangements to defeat the provisions of treaty, the onus must lie on the revenue authorities to establish the factual elements embedded in such allegations. It is only elementary that no one is expected to prove a negative. Hon’ble Supreme Court, in the case of K.P. Varghese v. Income-tax Officer [1981] 131 ITR 5971, has observed that “. . . to throw the burden of showing that there is no understatement of consideration, on the assessee, would be to cast an impossible burden upon him to establish a negative, namely, that he did not receive any consideration beyond that declared by him”. By the same analogy, an assessee cannot be expected to demonstrate that contracts are not artificially split, that the affairs are not so contrived so as to circumvent the duration test or that there is no maneuvering so as to abuse the treaty provisions. It is, therefore, for the revenue authorities to establish, beyond a reasonable degree of doubt, that there is an abuse of treaty provisions by so artificially contriving the affairs as to wrongfully entitle the assessee to treaty benefits. No doubt, in order to enable the revenue authorities to discharge this onus, the assessee must comply with reasonable requisitions of the revenue authorities and truthfully share the information available with him, but the exercise to establish treaty abuse is to be conducted by the revenue authorities. Unless that exercise is conducted, it cannot be open to disregard the claim of the assessee by simply making vague and generalized claims about artificial splitting of contracts and about the sham arrangements to defeat the treaty provisions. In the case before us, no such exercise is carried out. In the orders of the authorities below, a reference is made to the contracts having been awarded by one entity - directly or indirectly, and the fact that the work is carried out at the same place but these facts, by themselves, does not put the case in the category in which treaty provisions are abused by artificial arrangements, and, for that reason alone, the time spent on all the activities is required to be aggregated. The aggregation of time spent on various activities, on account of artificial splitting of contract by the assessee or other modes of treaty abuse, cannot, thus, apply unless the reasons embedded in this approach are established by the revenue. That is not the situation before us.

12. In our considered view, the only other situation in which aggregation of time spent of various activities is to be done is when the activities are so inextricably interconnected or interdependent that these are essentially required to be viewed as a coherent whole.

13. OECD Commentary has taken note of two such situations, i.e., (a) building site should be regarded as a single unit, even if it is based on several contracts, provided it forms a coherent whole commercially and geographically, and (b) when the very nature of construction or installation project may be such that the contractors activity is to be relocated continuously or at least from time to time (e.g., construction of roads and canals, dredging of waterways or laying of pipelines), the activities performed at each particular spot in a single project must be regarded as a single unit.

14. The test of ‘commercial and geographical coherence’, thus, does find a mention in the OECD Commentary though only with regard to a building site. Interestingly, this refers to such a degree of coherence that the different units, taken together, form a ‘coherent whole - geographically and commercially’. That cannot be equated with mere commercial and geographical coherence simplicitor in the normal course of business situations. The ambiguity of commercial and geographical coherence test apart, this test is not of universal application nor can it be construed as a conclusive test. There could be activities, such as construction of roads, which may or may not be geographically coherent but yet, according to the OECD Model Convention Commentary and in accordance with the fundamental rationale of construction PE concept, the time spent on progressive relocations is required to be aggregated. Similarly, there can be situations in which location of projects may be geographically the same, and yet as these are completely independent projects, the aggregation of time spent on the two projects may not be justified for that reason, as in the case of Sumitomo Corpn. v. Dy. CIT [2008] 114 ITD 61 (Delhi) where even though the situs of activities were at different parts of the same factory, viz., assembly floor, paint shop and weld shop”, yet the Tribunal came to the conclusion that “it cannot be said that all contracts put together formed a coherent whole-commercially or geographically”. On a conceptual note also, merely because different construction, project or supervisory activities are being carried out at nearby physical locations, these activities, for this reason alone, are not required to be seen in conjunction with each other. A construction site or project site inherently lacks permanence, since the construction, assembly, project or super-visory activity are supposed to continue for a limited time only, i.e., till the objective is achieved, and it is perhaps for this reason that the fictional PE for these types of activities is created so as to meet the situations when no PE taxation is triggered under the basic rule. This deeming fiction is to be applied for each construction or project site or supervisory activity in connection therewith. This deeming fiction, like all deeming fictions, is to be applied strictly. As an enterprise working in the other Contracting State, the situs of performing the activities, which triggers this fictional PE, is not necessarily a factor which is even controlled by the enterprise. It cannot, thus, have much bearing on the business model of the enterprise, and, therefore, on the question whether or not the enterprise is carrying on the business through the PE. As for the ‘commercial coherence’, there is hardly any consensus on its connotations either. Prof. Arvid Skaar, a well- known Norwegian International Tax Scholar, in his book “Permanent Establishment - Erosion of a Tax Treaty Principle - Third Indian Reprint, 2009, published by Kluwer Law International - at page 355” suggests that “it can normally be assumed that projects conducted under the same contract, i.e., the same document, will be considered a coherent whole” and expresses dissatisfaction about lack of clarity on the issue by adding that “apart from the assumption that one contract is one project, the identification rules of commentaries are sparse and obscure”. There are quite diversified opinions on the connotations of ‘coherent whole’ or ‘commercial coherence’. On one end of the spectrum, there is a decision of the Belgian Cour’ d Appel Anvers, 23 ET 387 [1983] referred to in Klaus Vogel on Double Taxation Conventions (Third Edition) - page 308 which seems to suggest that ‘the same ordering party’ will constitute commercial coherence, on the other end of the spectrum, there is also a decision by a Co-ordinate Bench of this Tribunal in the case of Sumitomo Corpn. (supra) wherein this theory is impliedly rejected and it is held that even when contracts are relating to different areas of manufacture of cars but these contracts are independent and not capable on bringing in a coherent whole, ‘mere commonality of principal cannot be sufficient’. The views expressed by the Tribunal are also on the same lines as expressed by Arvid A. Skaar in his book “Permanent Establishment - Erosion of a Tax Treaty Principle”. Third Indian Reprint, 2009, published by Kluwer Law Interna- tional - at page 365, wherein he has questioned the school of thought advocated by Klaus Vogel as also the Belgian ruling mentioned above, and observed as follows:

“The significance of identical clients has been particularly emphasized by German tax treaty commentators. To be a coherent whole, according to Vogel, the tasks have to be performed at the same place and for the same client. In the present author’s opinion, this position seems to presuppose that the creators of the commentaries agreed upon a specific criterion though they used an ambiguous one in the text of the commentaries. However, under treaty interpretation based on Vienna Convention, the intentions of the contracting parties are considered to be expressed in the treaty text. If the relatively precise criterion ‘identical client’ was intended, this would have been probably indicated by using a less ambiguous term than ‘commercially coherent whole’. . . .”

15. In the case of Sumitomo Corpn. (supra), as we have noted earlier, a co- ordinate Bench of this Tribunal has also impliedly rejected the emphasis on commonality of principal. While doing so, the Tribunal has, inter alia, observed as follows :

“79. Article 5(4) of Indo-Japan tax treaty [182 ITR (St.) 380] as was applicable in that case replaces the permanence element for existence of PE by the test of minimum length of time. In a case where there are several sites where supervision is going on in a country, the rule is that the test of minimum period should be determined for each individual site or installation project. Klaus Vogel in his commentary on Double Taxation Conventions, at page 308, has following to say on this aspect :

‘The question whether there is a PE in a specific contracting state or not should be considered separately for each activity performed in that State, i.e., for each individual place of business existing there as well. In this connection, the place where individual activity is performed may very well be relocated, for instance, where a road is being constructed in stages. If, in contrast, all building sites maintained in one State are treated as one single PE, this would in effect tantamount to force of attraction principle. Moreover, this would violate the principle that various business activities performed by one and same enterprise, none of which constitutes a PE, cannot lead to a PE, if combined.’

This above rule is, however, subject to exceptions, viz., where each building site or installation site forms a coherent whole in the other country and is operated at one place and by the same ordering party. The thrust of learned counsel for the revenue has been on this exception to the rule. We have already highlighted the fact that each purchase order was independent and did not complement each other. The MUL YE 2 project would not stand concluded with the execution of these purchase orders. The assessee was not the only person rendering these supervisory services. The sites were located at different places, viz., assembly floor, paint shop and weld shop. It cannot be said that all contracts put together formed a coherent whole - commercially or geographically. Even purchase orders relate to different areas of manufacture of car. . . . As already stated, perusal of purchase orders clearly indicate that the various contracts were independent and were not capable of bringing in a coherent whole commercially. Mere commonality of principal cannot be sufficient in this regard. . . .”

16. The two situations, referred to in the OECD Model Convention Commentary and which have been incorporated in UN Model Convention Commentary as well, are, thus, essentially illustrative in nature, and the common thread, and the highest common factor, in both these situations is that in both the cases, the activities are so inextricably interconnected that these cannot be viewed in isolation but only in conjunction with each other. The test of ‘geographical coherence’ and ‘commercial coherence’, in isolation with the larger picture of all the units forming part of a ‘coherent whole’, is not only a somewhat vague test with little consensus on its scope, and which can at best be loosely defined, but it is also somewhat unworkable in practical situations. In US Model Convention’s Technical Explanation - Technical Explanation to US Model Income Tax Convention - 1996, reference to “commercial and geographical coherence” is substituted by reference to the contracts or projects being “interdependent - both commercially and geographically”, and the said commentary, inter alia, states that “a series of contracts or projects by a contractor that are interdependent both commercially and geographically are to be treated as a single project for purposes of applying the twelve month (duration) threshold test”. The ‘interdependence’ test is something that can perhaps be applied with lesser ambiguity vis-a-vis ‘cohesion’ test simplicitor, and lesser ambiguity is certainly preferable. In any event, the highest common factor in both the examples set out in the OECD and UN Commentary is this ‘interdependence’ or ‘interconnection’. In view of the discussions above, we are of the considered view that the true test must lie in examining whether or not the activities performed by the enterprise in various projects or sites are interconnected and have to be necessarily regarded as a coherent whole. Unless the activities are of such a nature as to be viewed only in conjunction and as a coherent whole, in our humble understanding, there is no justification in aggregation of time spent on various business activities, sites or projects of the enterprise. In this view of the matter, strictly speaking, it is not really relevant whether the activities so carried out by the enterprise are for the same principal or different principals. The relevant considerations, in our considered view, are the nature of activities, their interconnection and interrelationship and whether these activities are required to be essentially regarded as a coherent whole in conjunction with each other.

17. It is, thus, clear that the justification for aggregation of time spent by the assessee on different project sites, for applying threshold of duration test, is not sustainable. Neither the work having been carried out for the same principal is sufficient to justify the aggregation of time spent on all the projects, nor the fact that this work was carried out in the same oilfield, which is a huge geographical area anyway, is sufficient to invoke that exercise. Even if these projects are commercially coherent in the sense that these projects are for the same principal, and geographically coherent in the sense that these are on nearby locations, these two factors would not necessarily mean that these projects are to be necessarily seen as a coherent whole - geographically and commercially. The true test, as we have noted above, is in interconnection and independence - in addition to geographical proximity and commercial nexus. There is no finding, nor even a suggestion, by any of the authorities below to the effect that the three contracts are inextricably interconnected, interdependent or can only be seen only as a coherent whole in conjunction with each other. The very foundation of the approach adopted by the CIT(A) is devoid of any legally sustainable merits. For all these reasons, we are unable to approve the findings of the authorities below in this regard.

18. We have, however, noted that the Assessing Officer has given a finding that, even on standalone basis, one of the contracts exceeds the threshold limit. It is pointed out that so far as Contract No. 95/C/04 (D-4522) is concerned, the actual contract period is concerned the same is nineteen months since the first invoice was raised on 1-11-1996 and the last invoice was raised on 4-5-1998. It is, however, noted that the commencement date, as per the contract, was 1-11-1996 and the completion date of the contract was specified as 31-5-1997. In appeal, the assessee had pointed out to the CIT(A) the invoices in question were for mobilization advance (invoice dated 1-11-1996) and for sail out of the Derrick Barge from outside India (invoice dated 4-5-1998), which have nothing to do with the actual work carried out at the site. It was also submitted that the actual work, as per completion certificates which were filed during the assessment proceedings, was started on 5-2-1997 and ended on 31-5-1997. The CIT(A), however, did not adjudicate on these factual issues as she concluded that the total time spent on all the contracts in India put together exceeds nine months, and, for that reason alone, the assessee could be said to have a PE in India.

19. Having held that the aggregation of time on various contracts in India is not required, this is nevertheless to be examined whether time spent on a specific contract is more than nine months or not. The CIT(A) ought to have examined this aspect of the matter and given her findings on duration of each project site. We, therefore, consider it necessary to remit the matter for the limited purposes of adjudication on this aspect of the matter. Even as we do so, it is necessary to point out that what is required to be ascertained is the time spent on the contract and it is nothing to do with the dates on which invoices are raised for advances and for demobilizations and sail outs. As the UN Model Convention Commentary, incorporating OECD Model Convention Commentary, puts it, “a site exists from the date on which the contractor begins his work, including any preparatory work, in the country where the construction is to be established, e.g., if he installs a planning office for the construction” and that “in general, it continues to exist until the work is completed or permanently abandoned.” These commentaries further add that “a site should not be regarded as ceasing to exist when work is temporarily discontinued.” These observations very appropriately set out the right principles in the light of which the duration of a site is to be ascertained for the purpose of applying threshold limit of duration test. The date on which invoice is raised for mobilization advance, as adopted by the Assessing Officer in this case, is also not decisive for the purpose of determining the duration of the PE. On what date a requisition is made for advance has nothing to do with the actual work, or incidental preparatory work preceding, and, thus, forming integral part of, the actual core work, at site. Similarly, sail out of barge is an activity which takes place after the work at site comes to an end. The date on which sail out of barge starts or is completed is essentially a date subsequent to abandoning the work at the site. This date is also not, therefore, relevant for the purpose of deciding the date on which activity at site comes to an end. It is also important to bear in mind the fact that dates of commencement and completion of work set out in the contracts are only indicative of plans and cannot be substituted for the actual dates of commencement of work and completion of work - as evidenced by the material on record. The work schedule set out in the contract cannot, therefore, be decisive of the date of commencement and completion of work at site. What is really required to be seen, as we have elaborated above, is the duration for which the work-actual or preparatary, is carried out at site.

20. While deciding the matter afresh on the question of duration of project, the CIT(A) will bear in mind the principles set out above and shall decide the matter by way of a speaking order dealing with the contentions of the assessee. With these directions and in the terms indicated above, the matter stands restored to the file of the CIT(A).

21. In the result, the appeal is allowed for statistical purposes in the terms indicated above.