The author contends that BALCO's finding on the seat courts having jurisdiction even if they wouldn't have had jurisdiction under section 16-20 CPC (i.e. "neutral venue"), is itself erroneous. However, Para 96 of BALCO does not say that ONLY the seat court would have jurisdiction; it says that both, the "neutral venue"/seat court AND courts under section 16-20 CPC would have jurisdiction. This continues to be the binding position of law. The major error committed in BGS Soma is misinterpretation of Para 96 of BALCO to say that ONLY the seat court would have jurisdiction and therefore not following BALCO which is binding and holds the field.
The Arbitration and Conciliation Act, 1996 ("the 1996 Act") was enacted in order to meet the challenges of the time, fix lacunae in The Arbitration Act, 1940 ("the 1940 Act") and promote India as a commercial destination. The Preamble of the 1996 Act notes that the UNCITRAL published a Model Law on International Commercial Arbitration in 1985 and the UN recommended that all member countries should attempt to bring their national laws in line with the same as far as possible. It is important to note that India did not adopt the Model Law wholesale, but merely took the same into account, as mentioned in various judgments. Therefore, one cannot assume that all the principles of the Model Law are applicable in Indian law.
"Court", "Seat" and BALCO:
The definition of "court" under Section 2(1)(c) of the 1940 Act at its heart, was substantially the same as the present definition under Section 2(1)(e) of the1996 Act. Apart from the first line identifying the kind of court, both contain similar words: while the 1996 Act says "… having jurisdiction to decide the questions forming subject-matter of the arbitration if the same had been subject matter of a suit…", the 1940 Act says "having jurisdiction to decide the questions forming subject-matter of a suit". In 1971, the Supreme Court in Hakam Singh v. Gammon India  held in a case under the 1940 Act, that even by consent, parties could not confer jurisdiction on a place that otherwise would not have had it under Section 16-20 of the Code of Civil Procedure 1908 ("CPC"). This was followed in ABC Laminart v. A.P. Agencies, Salem and has been settled law ever since. The reason for this of course is that the definition of "court" in the 1940 Act clearly gave jurisdiction akin to that for a civil suit, i.e. based on subject matter.
A 3- Judge bench of the Supreme Court in Bhatia International v. Bulk Trading S.A. held that Part I of the 1996 Act is applicable to international commercial arbitrations held in India and also those outside India. This was reversed by a 5-Judge bench in Bharat Aluminium Co. v. Kaiser Aluminium Technical Services Inc. ("BALCO") which said that Part I will apply only to arbitrations seated in India. The concept of seat vs venue was also explained in terms of Section 20 of the 1996 Act. This judgment has been followed and interpreted in numerous judgments since then and holds the field even today.
Significantly, BALCO (supra) was the first judgment to authoritatively say at Para 96 that even if the parties by agreement had chosen a "neutral venue" as the seat of the arbitration, the courts of the seat would have jurisdiction. The same was more clearly stated later in Indus Mobile Distribution Private Limited v. Datawind Innovations Private Limited. Before analysing what else the problematic Para 96 really said or meant, it's important to note that even this was a departure from the settled law which created a new principle applicable specially to arbitrations, on the ground that else Section 20 of the 1996 Act, which allows party autonomy, would become redundant. In the opinion of the author it would not, since Section 20 would still allow parties to seat the arbitration wherever they wish, just that the mere fact of a place being the seat would not convert it from being a "neutral venue" to a place whose courts would have jurisdiction. In the opinion of the author, though this would be correct for arbitrations seated outside India, there was no warrant in law for reaching this conclusion with regard to arbitrations held in India. This is because a supervisory "court" could only be as defined in Section 2(1)(e) of the 1996 Act, which necessarily brings in Sections 16-20 of the CPC since for the purpose of identifying this "court", one has to treat the "subject matter of the arbitration" as if it were "subject matter of a suit". In the humble opinion of the author, this finding needs to be re-examined by a bench of 7 Hon'ble Supreme Court judges.
The Judgment in BGS Soma:
BGS SGS Soma JV vs NHPC Ltd.  ("Soma") was decided by a 3 Judge bench of the Supreme Court which was bound by the 5-judge bench's decision in BALCO as it stands. Soma was a case of a purely domestic arbitration dispute where both parties were Indian and the seat designated by the agreement was "New Delhi/Faridabad". A summary of what Soma holds is:
Analysis of the reasoning in Soma
Misplaced reliance on judgments:
Soma relies on numerous Indian as well as English judgments on the subject of deciding the seat, seat versus venue and also the question of which country's courts would have supervisory jurisdiction. Every single one of the English Judgments referred to in Soma, without exception, are those in which the contest is between courts of country A versus those of country B. Not even a single English case deals with a domestic dispute, one where the contest is between two places in the same country. Therefore, reliance on those English judgments, it is respectfully submitted, was factually misplaced. This is quite apart from the fact that Indian law is not at all the same as English law or the UNCITRAL rules. Furthermore, except for Indus Mobile (supra), all the Indian judgments discussed by Soma (supra), were also those in which the contest is between courts of one country versus those of another country. It is submitted that the judgments relied upon and reasoning given by the bench in Indus Mobile (supra) are exactly the same as that of Soma (supra) and for that reason it is respectfully submitted that in the opinion of the author, even Indus Mobile (supra) has been incorrectly decided.
Erroneous interpretation of Para 96 of BALCO:
BALCO (and all the subsequent judgments), was a case of international commercial arbitration where the clause provided for arbitration "wholly in London". The heart of the case was the interpretation of Section 2(2) of the 1996 Act and the omission of the word "only" from it, when compared to Article 1(2) of the UNCITRAL Model Law. Appreciating this is crucial and it is why the Court ultimately held that the said omission would make no difference and that Part I would apply only to arbitrations seated in India. Therefore, the entire discussion and ratio in BALCO (including and particularly Para 96) as well as all the other judgments relied upon in Soma, should be looked at in the light of what they were considering, viz. a contest between supervisory jurisdiction of courts of the country of the seat versus those of the subject-matter country. It should not be taken as applicable to a purely domestic arbitration where the location of the seat as well as the subject-matter/cause of action are both Indian.
Now for Para 96 of BALCO (supra) itself. At placitum 'b' of the SCC report it is stated that "in our view, the legislature has intentionally given jurisdiction to two courts, i.e. the court which would have jurisdiction where the cause of action is located and the courts where the arbitration takes place". Even the final sentence says that both courts would have jurisdiction. However, the confusion arises when it also says in the example and the penultimate sentence, that if the seat is Delhi and Delhi is a neutral place, then "… appeal against such an interim order under Section 37 must lie to the courts at Delhi…". This has been interpreted in Indus Mobile, Soma and subsequent judgments to mean that only Delhi courts or courts of the seat can entertain a Section 37 appeal against an order passed under section 17. In the respectful opinion of the author, this is an incorrect interpretation, as what Para 96 really says, is that the appeal under Section 37 "must lie" i.e. the appeal "does lie" or "also lies" to Delhi courts as well as Mumbai and Kolkata - not that it lies to Delhi courts exclusively. It is respectfully submitted that only if this interpretation is taken, does the wholistic reading of para 96 make sense and is consistent. The interpretation in Soma (supra) and Indus Mobile (supra) leads to the incongruous result that subject-matter courts, which have expressly been given jurisdiction as per the definition of "Court", would be denuded of it. It amounts to a complete judicial reversal of what the Legislature expressly provides for Indian seated arbitrations in the 1996 Act in 2 stages. First, by BALCO ALSO allowing neutral (seat) courts to exercise jurisdiction de hors Section 16-20, CPC and second, Soma going a step further and saying that ONLY the seat courts will have jurisdiction. This is done by relying on purposive interpretation when the literal interpretation is clear and by relying on judgments decided by foreign courts in a different, non-domestic context.
Curious interpretation of Section 42:
Antrix at Para 59 said that if only seat courts are held to have jurisdiction, it would make Section 42 otiose. Soma at Para 62 says this is incorrect (discussed below). It says that obviously, any application under Part I (Section 9, 11, 34 etc) has to only be made to a court that has jurisdiction, which can only be the seat court and as per Section 42, all subsequent applications must be made in that court only. Further, in a situation where no seat is indicated in the agreement and before the Tribunal decides a seat under Section 20(2), a party can file an application under Section 9 or 11 in any jurisdictional court. All subsequent applications then can only be filed in that court as per section 42- for both these reasons, Section 42 is not redundant.
The author submits that firstly the plain wording of Section 42 of the 1996 Act does not remotely lend itself to such an interpretation. Secondly, it would mean that only the court of first filing would have jurisdiction for all time, thereby making any determination of seat otiose. It clearly goes against the theory of primacy of the seat and the seat court having exclusive jurisdiction, that by mere strategy of a cunning party, the seat courts can be completely avoided. Therefore, it is humbly submitted that neither this reasoning nor the interpretation that leads to it is correct and breaks down when logically tested.
The conclusion in Soma:
Lastly, Soma (supra) concluded at para 99-101 on its facts, that applying the Shashoua principle, either Delhi or Faridabad can be the seat- whether the dispute is of NHPC with an Indian contractor or a foreign contractor. Even the other language suggests that New Delhi/Faridabad is not merely the venue but is the seat. It further held that, however, since proceedings were finally held at New Delhi and the award was signed at Delhi, it shows that the parties had agreed that New Delhi would be the seat under section 20(1). Therefore, even though a part of cause of action may have arisen at Faridabad, it will not have jurisdiction once the seat has been determined.
It is respectfully submitted that this finding is incorrect as the same is clearly contrary to Videocon Industries Limited v. UOI , on which Soma itself relies and which holds that the seat once fixed cannot be changed except by writing signed by the parties. It's only the venue that changes. Therefore, simply holding sittings at Delhi should not change the seat if it was actually Faridabad, or if both can be the seat then both should remain the seat. Furthermore, between the two places, some cause of action had arisen at Faridabad since the Agreement was signed at Faridabad and notices were sent by Petitioner to the Respondent's registered office at Faridabad. The only thing that happened at Delhi was the holding of arbitral hearing and the award.
No exception can be taken to the judgment in Soma in so far as its application of the Shashoua principle to foreign seated arbitrations goes, or to its finding of Hardy Exploration (supra) to be contrary to the same principle as approved in BALCO. The reasoning and interpretation in Soma is no doubt well-intentioned but unfortunately, and with great respect it is submitted that the judgment states the law as it SHOULD be and not what the law actually IS with respect to an Indian seated arbitration.
 (1971) 1 SCC 286
 (1989) 2 SCC 163
 (2002) 4 SCC 105
 (2012) 9 SCC 552
 (2017) 7 SCC 678
 2019 SCC OnLine SC 1585
 2018 SCC OnLine Del 9338
 (2019) 13 SCC 472
 (2011) 6 SCC 161