VMZB 2021: the new insurance exchange conditions for fire insurance
Author: Patrick van der Vorst, attorney-at-law and partner in VanNiekerkCieremans Advocaten in Rotterdam, and editor of the online law journal VAST. Date of publication: 22 November 2021. This article is a translation of the Dutch article published in the online law journal VAST. See: P.R. van der Vorst, "De nieuwe beursbrandvoorwaarden: de VMZB 2021", M.A.D. Lex: VAST 2021, P-053.
On 28 June of this year, the Netherlands Insurance Exchange Association (VNAB) presented its new model terms and conditions for fire insurance: the VNAB Modular Conditions for Property Damage and Loss of Profits Insurance 2021 (Dutch abbreviation: VMZB 2021). This is a model or basic policy which provides cover for both property damage and loss of profits (business interruption loss). As the successor to the widely used NBZB 2006 Dutch Insurance Exchange Conditions for Property Damage and Loss of Profits Insurance, these new model conditions can be seen as an evolution of their now 15-year-old predecessor.
1. Introduction
In this article, I will first consider the history of the insurance exchange fire conditions and then discuss the system and content of the VMZB 2021. For the purpose of this discussion, I will proceed on the premise that the interpretation of insurance exchange conditions depends primarily on objective factors, such as the wording of the provision when read in the context of the policy conditions as a whole and any accompanying explanatory notes.[1] The reason for considering the history of the VMZB 2021 conditions is that it is apparent from the published explanatory notes to them that that their content largely matches that of the NBZB 2006 conditions. The interpretation of the new conditions is therefore based on that of the NBZB 2006.
This article is intended to provide the reader with an overview[2] by describing in a broader context the nature and operation of the VMZB 2021. As I cannot deal with the subject exhaustively here, I propose instead to concentrate on a few main aspects of the substantive cover and, where necessary, indicate how previous conditions may be relevant to understanding and applying the VMZB 2021. Finally, I will discuss some changes to the general conditions and definitions.
2. History in a nutshell
The VMZB 2021 conditions, like the NBZB 2006, are a policy that offers combined coverage for property damage and loss of profits. This combination is not self-evident. The VMZB 2021 are based on separate model conditions which were later combined. The ‘property damage module’ dates back to 1998. That year saw the introduction of the Dutch Insurance Exchange Policy for Comprehensive Perils (NBUG 1998), together with Explanatory Notes.
In contrast to their predecessor (the NBBP 1990), the NBUG 1998 conditions listed 25 perils for which coverage was intended to be provided. According to the Explanatory Notes, those perils (but also the ‘extras’ included under the policy) ‘broadly corresponded (…) to what was customary in the market at that time in varying forms’. The VMZB 2021 include these 25 perils in the list of covered perils in almost identical terms. In this respect, little has changed and the NBUG 1998 can be used as a source of inspiration when interpreting the VMZB 2021. As the Explanatory Notes to the NBUG 1998 contain background information on the various definitions and descriptions of coverage used in the NBUG 1998, they remain relevant in principle.
Subsequently, in the autumn of 2001, the NBUG 2002 conditions were introduced to rectify certain ‘beauty flaws’ in the 1998 text and to adapt it to allow for the introduction of the euro.
The history of the loss of profits model conditions has followed a more or less similar course: the first version appeared in 1999 and the VNAB introduced the Dutch Insurance Exchange Loss of Profits Policy for Comprehensive Perils (NBBU 2002) in the autumn of 2001. In terms of style and coverage, these conditions mirrored the NBUG and they too were published with Explanatory Notes.
That moment was immediately seized upon to introduce combined property damage and loss of profits cover: the Dutch Insurance Exchange Conditions for Property Damage and Loss of Profits Insurance (NBZB 2002).
In late 2005, the 2002 versions of the NBUG, NBBU and NBZB were updated to take account of the introduction of new insurance legislation. These updated texts were the 2006 editions of the NBUG, NBBU and NBZB conditions.
The VMZB 2021 are based on those conditions, more specifically on the NBZB 2006. Although the VMZB 2021 conditions are not identical in terms of intended cover to the NBZB 2006 currently used in the market, the intention is that the VMZB 2021 should largely correspond to them. See the Explanatory Notes to the VMZB 2021 published on 28 June 2021 (referred to below as: ‘Explanatory Notes to the VMZB 2021’): ‘The VMZB 2021 have been developed by the VNAB to better reflect the current insurance market, albeit without making major changes to the coverage.’ (page 1 of the Explanatory Notes to the VMZB 2021).
For practical purposes, it is important to note that the Explanatory Notes to the VMZB 2021 shed light on the differences between the new conditions and the NBZB 2006 in terms of both substantive changes and linguistic clarifications. A 41-page document comparing the NBZB 2006 and the VMZB 2021 provides additional insight, with the differences in topics being grouped in tabular form. The VNAB website also provides answers to some questions asked during an online presentation about the VMZB 2021.[3]
The Explanatory Notes to the VMZB 2021 and the comparison version do not provide an exhaustive explanation of the operation and content of the policy coverage under the VMZB 2021. The Explanatory Notes to the old NBUG and NBBU conditions and the relevant case law remain important for the interpretation of policies to which the VMZB 2021 are declared applicable. Naturally, the views of the insurance exchange on the interpretation of certain terms also remain important. After all, as the Explanatory Notes to the VMZB 2021 state that the VMZB 2021 are in line with the ‘current insurance market’, it is apparent that no derogation is intended from what the exchange currently understands by certain terms/coverage under the NBZB 2006 regime. For these reasons, it is right to continue consulting the VMZB 2021’s predecessors.
3. VMBZ 2021 system
As already noted, the Explanatory Notes to the VMZB 2021 do not tell the whole story. For example, they do not describe the nature and system of the coverage, as the drafters assumed that this would be known to the insurance market. That is why I will touch on this (briefly) in this article.
Like their predecessors, the VMZB 2021 conditions are based on a system of named perils: in other words, damage to insured risk objects is covered only if that damage is caused by a peril specified in the policy (article 1, first sentence, Module I). If the peril that causes damage is not specified in the policy, there is no coverage. For example, damage due to corrosion,[4] design errors or subsidence/ sinkholes[5] is not named as an insured peril, nor is damage due to water accumulation.[6] If an unnamed peril independently causes property damage, that damage is not covered by the policy. However, if the parties to the insurance contract do wish to insure against such perils and other non-standard insured perils, they must arrange for this to be specified in a separate clause.
But what is the position if the insured fire risk results in an unnamed (and hence uninsured) peril and damage results from that uninsured peril? For such situations, article 1, second sentence of Module I provides that such damage is covered if the uninsured cause is the direct result of an insured peril. In short, this provision can be regarded as an ensuing loss clause, in which a chain of separate perils results in damage (for example, where a toxic cloud is released by a fire (see the Explanatory Notes to the NBUG 2002)). As this is relevant only very occasionally, I will confine myself for the purposes of this article to pointing out that this coverage (which has the pitfall that parties can become embroiled in a semantic debate) has been adopted unchanged from the NBZB 2006.
The principle underlying the system of a named perils policy is that only named perils are insured. The unnamed perils are not covered and do not need to be excluded by means of an exclusion clause. This is because the parties indicate, by concluding an insurance contract based on named perils, that coverage exists only for damage caused by those named perils. If those positive conditions of the primary coverage description are not met, the policy does not provide coverage. This means that an extensive list of exclusions is not necessary in the case of a named perils policy. Nonetheless, exclusions are still sometimes included in order to avoid any misunderstanding about the fact that damage is not covered.
As regards the exclusions in the case of the VMZB 2021, it is worth noting that the nature of the coverage provided under those conditions should not be lost sight of. In the case of fire damage, it should be remembered that the policy provides coverage against the risk of fire, regardless of the cause of the fire. This could be a short-circuit in a machine, but the fire could equally well have been caused by an unknown (dominant) factor or by a form of uninsurable war risk or civil unrest (acts of war). In the latter case, the exclusion for war risks and civil unrest included in the VMZB 2021 (article 4.1 of Module I) means that a building burned down due to rioting would be excluded from coverage.
Policies based on an all-risks system do contain an extensive list of exclusions. In principle, such a policy covers all risks, but this broad coverage is then trimmed back by an extensive list of exceptions. It might be assumed from the description of a policy as ‘all risks’ that all risks really do fall under the primary coverage clause. This seems to be a fairly common misconception, particularly among people outside the insurance market. In my view, the label ‘all risks’ mainly serves to classify the system on which the policy is based. It is clear, after all, from the text of the positive coverage clause itself that not all risks are covered simply as a matter of course, as this states that coverage exists only for risks that are of an ‘ongevalsmatig karakter’ (in the nature of an accident). This is often done by including the phrase ‘plotseling en onvoorzien’[7] (sudden and unforeseen) in the primary coverage description, although the term ‘onvoorzien’ (unforeseen) is sometimes replaced by ‘onzeker’ (uncertain), ‘accidenteel’ (accidental) or ‘onverwacht’ (unexpected) (see also the terms used in English versions: ‘sudden and accidental’, ‘sudden and unforeseen’ etc.).
The advantage of an all-risks policy is that all exclusions are neatly listed, whereas in a named perils system they tend to be more implicit. This can be less transparent for policyholders who are not insurance professionals. However, in professional insurance practice this is a tried and tested approach.[8] As the knowledge and expertise of the insurance broker must, in principle, be attributed to the insured, this does not result in any significant problems in practice.
This concludes my general observations about the system of the VMZB 2021 conditions and how they differ from other types of policies. It would be beyond the scope of this article to delve more deeply into the various types of policies used in the market, including hybrid variants, or to consider (possible) differences in effect in terms of burden of proof and causality.
4. The content of the VMZB 2021
Introduction
The VMBZ 2021 conditions consist of four sections: 1) a section containing general policy conditions (p. 4 et seq.), 2) a property damage module (p. 15 et seq.), 3) a loss of profits module (p. 28 et seq.) and 4) a section containing definitions (p. 33 et seq.). The policy schedule specifies which module or modules apply and what the insured sums and deductibles are. Below, I will briefly discuss the content.
Module I – property damage
The property damage module provides coverage against damage to or loss of insured risk objects to the extent that this is caused by the named perils/occurrences. These are, broadly speaking, the same 25 perils/occurrences that were already covered under the NBZB 2006. As important coverage for perils such as fire and storm has remained unchanged, the interpretation of these provisions can be based on how this coverage was (and still is) applied under the NBZB 2006, its predecessors and similar conditions.
Reference can continue to be made to the old conditions in the case of causality issues, as the VMZB 2021 make no change in this respect. However, following the Supreme Court's judgment of 4 June 2021,[9] it may be wondered in this connection whether the parties would not be wise to specify in the policy what standard of causality (e.g. the theory of dominant cause, attribution according to reasonableness, causa remota, etc.) applies. Naturally, for the sake of certainty everything can be nailed down to the last detail, but, in my view, the history of the VMZB 2021 and their use in the insurance exchange context mean, in brief, that there is no obvious need for such an approach.
As various parts of the coverage for water/precipitation damage have been changed in the VMZB 2021, I will discuss this subject in more detail in this article. Whereas the NBZB 2006 conditions contain an exclusion in article 2.2.8.5 for damage caused by (i) structural defects and (ii) poor maintenance of the building, the scope of this exclusion is seemingly limited in article 2.8 of the VMZB 2021 by a number of textual additions, which makes the exclusion seem less broad. For example, it now follows from the literal text that water damage caused by structural defects is excluded only if the construction errors occurred during the ‘construction, extension, alteration or renovation’ and only to the extent that this is ‘within the sphere of influence of the insured and/or companies related to the insured’. The latter also applies to poor maintenance of the building.
It is notable that the Explanatory Notes to the VMZB 2021 (p. 3) describe these additions as merely clarifications of the interpretation of the terms ‘structural defect’ and ‘poor maintenance’ used in the NBZB 2006. In my view, that is not the case. To my knowledge, there is nothing to suggest that the scope restrictions now incorporated in the VMZB 2021 were (already) deemed applicable under the NBZB 2006 conditions.[10] In my view, these are new elements which (compared with not only the NBZB 2006 but also the NBUG 1998) impose additional requirements on the applicability of the exclusion under the operation of the VMZB 2021.
As regards the term structural defect, the Explanatory Notes to the VMZB 2021 state that the exclusion applies if the insured is the client, i.e. the party commissioning the construction, extension, alteration or renovation, and a structural defect is caused by the contractor while carrying out such work. According to the Explanatory Notes, the exclusion does not apply if the insured is merely the tenant of the insured building or simply the purchaser of an existing building that already has such a defect. According to the Explanatory Notes, the decisive factor would therefore seem to be the capacity of the insured in relation to the work performed. As this is a new element, I will consider it in more detail below.
Upon first reading of the Explanatory Notes, the underlying rationale for the addition to the exclusion clause is not immediately apparent. After all, this is an exclusion for structural defects for which the insured itself does not incur any blame. In the example given of a contractor whose error results in a structural defect, a client/insured will often be unaware that the building has had a structural defect from the outset. Nevertheless, under the policy this defect is attributed to the insured because it is ‘within its sphere of influence’ since the insured is the party commissioning the contractor. This seems to introduce an element of risk criterion. If such defects are considered to be within the insured’s sphere of influence, what about other relationships in which the risk of errors causing structural defects is borne contractually by the insured? I am thinking, for example, of cases involving the purchase of an existing building in which the purchaser is made fully liable for such risks or in which the insured was (or should have been) aware of the defect in the construction of its building.
As the Explanatory Notes to the VMZB 2021 (p. 4) state that, in the case of the exclusion for poor maintenance (to which the same phrase ‘within the sphere of influence of the insured’ applies), the decisive factor is deemed to be which party is contractually liable, I see no reason why this should be different in the case of structural defects. For example, as regards poor maintenance the Explanatory Notes merely state:
‘Also, the inclusion of the words ‘within the sphere of influence of the insured…’ in this clause provides a clear indication of how the poor maintenance exclusion should be interpreted. In principle, the basic assumption here is that the building owner is responsible for the maintenance. Naturally, different arrangements can be made in tenancy agreements or other documents, or special circumstances can play a role. The interpretation of this exclusion must therefore be assessed on a case-by-case basis. [Underlining added].
In the case of structural defects, this seems to provide scope for other situations where the exclusion applies – i.e. situations other than those currently mentioned in the Explanatory Notes. Once again, it should be noted that in comparison with the old conditions the addition in the case of poor maintenance is (also) new. According to the Explanatory Note on poor maintenance quoted above, how the risk is attributed contractually is decisive in determining whether the exclusion applies. As this applies in the case of poor maintenance and no textual distinction is made as regards structural defects, it could be argued that who is (ultimately) liable is just as important in the case of structural defects and also that ‘special circumstances’ can always play a role in this.
In short, this addition in the VMZB 2021 has not made things any easier. Whereas under the NBZB 2006 the question of who is liable for the structural defect or poor maintenance (which, as an independent ‘event’ is not an insured risk at all) is not relevant, this does seem to play a role in the case of this provision (according to the literal text of the Explanatory Notes, in any event in the case of poor maintenance).
Naturally, the parties are not bound to accept the text as currently worded: the VMZB 2021, like its predecessors, serves as a model from which the parties can derogate. In view of the foregoing, it seems to me that it would be advisable to clarify how the provision should be interpreted (unless the parties simply choose to fall back on the ‘old’ wording). Nor is the objectively recognisable intention of the exclusion clear from the rest of the Explanatory Notes either. On the one hand, the impression is created that, in substance, the coverage under the VMZB 2021 has remained the same as under the NBZB 2006. On the other, in the case of this exclusion, the ‘clarification’ that has been inserted in the text itself actually seems to introduce a new element rather than clarify. These interpretations clash with one another and raise the fundamental question of whether or not the (old) case law and the manner in which market participants have for many years interpreted this provision of the NBZB 2006 still apply.
What, in my view, is beyond discussion (although the Explanatory Notes do not say as much explicitly) is that the exclusion for structural defects includes not only design errors (such as the example given in the Explanatory Notes of the contractor that makes an error) but also all errors in the construction (regardless of the cause). In my view, defects in the construction caused by execution errors (including delivery of defective materials) also fall under the concept of structural defects. In a case concerning the NBUG 2002, Amsterdam Court of Appeal held as follows on this point:
‘As the District Court has held on good grounds, which the Court of Appeal hereby adopts and makes its own, it is important that the term ‘structural defect’ is included in the text of the relevant policy clause without further specification. In common parlance, the term ‘structural defect’ refers not only to design errors but also to errors in the construction as a result of execution errors. It must therefore be concluded that there is no reason for the limited interpretation advocated by Remco et al. and Drienerveld. Moreover, the interpretation of the term ‘structural defect’ advocated by Remco et al. is not compatible with the fact that the clause also excludes damage caused by poor maintenance of the building. As the insurers have rightly argued, this shows that they did not wish to provide cover for damage caused by defects in the building, whether these occurred during the designing of the building, during the implementation of the design or, after construction, during the maintenance of the building.’[11]
In my opinion, this case law remains as important as ever when interpreting the VMZB 2021.
As previously noted, the case law also retains its relevance for the interpretation of other terms used in the VMZB 2021. In this context, it is worth mentioning that the requirement that the defect should occur ‘suddenly’ as referred to in article 2.8.1 of the NBZB 2006 (now article 2.8 (a) of the VMZB 2021) often leads in practice to debate where damage is caused by water leaking from pipes. The text of the policy requires that the defect must have occurred ‘suddenly’. How that term is interpreted gives rise to differences of opinion, just as with the application in practice of the word ‘sudden’ in all risk policies. For example, there may be discussion about whether a hose coming loose and causing water damage qualifies as a ‘sudden’ defect if this is due to a clamp not being properly tightened. The introduction of the VMZB 2021 does not aim to settle the matter once and for all. A relevant factor that may play a role in the interpretation is perhaps the explanatory note provided by the VNAB’s Technical Fire Committee in this connection during the online presentation of the VMZB 2021. For example, one of the subjects dealt with during that presentation was whether damage caused by the fact that a third party had not properly tightened a clamp was covered by the article. The answer given by the Technical Fire Committee was that a defect could not be said to have occurred suddenly if something had not been fitted or functioned properly.[12] Quite apart from the question of whether statements of this kind made during a policy presentation have any legal value, the committee in question prefaced their answer with a disclaimer to the effect that this was ‘fodder for lawyers’ and things might well turn out differently in court.
In my view, however, there is nothing strange in the idea that something that has always been broken/defective cannot ‘suddenly’ become defective at a later stage. After all, it was already broken or malfunctioning. This is in keeping with the prevailing view in the literature and case law on the doctrine of material damage, namely that a thing must first be ‘in good condition’ before it can be damaged.
Module II – loss of profits
The coverage provision for loss of profits in the VMZB 2021 adopts the same system as in the NBZB 2006: in so far as property damage is insured under Module I, any resulting loss of profits is also covered. What is new in comparison with the NBZB 2006 is that there is now a choice between ways of settling the claim, namely on the basis of the reduction in either (a) turnover or (b) production. It is important to note that the insured must make this choice by notifying the insurers’ loss adjusters before the start of the loss adjustment.
As a reduction in turnover is not the same as a reduction in production, the decision by an insured to choose one or the other way of assessing the extent of the loss of profits covered under the policy may turn out to be advantageous or disadvantageous. An insured (assisted by its broker) would therefore be well advised to consider in advance how this could possibly turn out for its business.
From the insurer’s point of view too, it is worthwhile taking this into account. If recourse for the loss exists against a liable party, it may be the case that production loss paid out under the policy cannot simply be recovered from that party. This is because a production loss does not necessarily entail a turnover loss that can be recovered from a liable party,[13] for example where there is still sufficient stock available.
General conditions section and definitions section
The section of the VMZB 2021 containing the general conditions makes several changes to the provisions of the NBZB 2006. Four of them are reviewed below.
First, carrying on from the last topic in the previous part, article 11 provides that all insurers involved in the policy must comply with the Fire Insurance (Right of Recourse) Sectoral Regulations published by the Dutch Association of Insurers if they seek recourse. This has been done so that insurers that are not members of the Association are also bound by those regulations. For these non-member insurers, the unconditional applicability of the VMZB 2021 to the policy means that their recourse options are restricted. The main restrictions are, in brief, that (i) recourse is possible only if the liable party can (also) be held liable for the unlawful act/omission of a person and (ii) no recourse is possible against (a) private individuals and (b) commercial and other non-private tenants, agricultural tenants, lessees, borrowers and custodians of the damaged property, and (iii) in principle, third parties seeking recourse have priority over insurers for amounts to be recovered.
Second, the maximum level of the expenses of the insured’s loss assessor eligible for reimbursement has been raised from the level specified in the NBZB 2006. Whereas, under article 8.4 of the NBZB 2006, this was still limited to a maximum of the fees and expenses of the loss adjusters appointed by insurers, this has now been increased to a maximum of 130% of this amount in article 5 of the VNZB 2021.
Third, it should be noted that under article 24 of the VMZB 2021 any disputes may now only be submitted to Rotterdam District Court. Submission to Amsterdam District Court is no longer possible.
Finally, it should be noted that the definition of ‘building’ has been amended/expanded. Under the old conditions, there was sometimes debate as to whether or not a structure should be classified as a ‘building’.[14] Now, the definition of building has been altered to include: ‘(...) any sizeable structure made of wood, stone, metal or other material. Such a structure must be directly or indirectly connected to the ground, or have its supports in or on the ground, such as but not limited to sheds, warehouses, portacabins and fences.’ Moreover, the Explanatory Notes to the VMZB 2021 (p. 6) state that solar panel systems are treated as part of the building if they are connected to it (i.e. roof-mounted solar panels), but that stand-alone solar panel systems (located next to the building) ‘are not a building within this definition’. The drafters apparently intend that the panels must then be insured separately and cannot therefore automatically be included in the coverage. According to the FAQ on the VNAB website, ‘equipment installed on a parking area on the site is covered under the insurance, provided that it is connected to the building’. Where the situation is unclear, it makes sense to specify this in a separate clause in individual cases.
5. Conclusion
The VMZB 2021 conditions are, for the most part, a continuation of the widely used NBZB 2006. It follows that the case law, literature and opinions relating to the 2006 conditions remain largely valid. Nonetheless, parties would be well advised to take account of the points that have changed – or seemingly changed – as a result of the introduction of the VMBZ 2021. Some of these changes, the most important of which I have touched upon in this article, may have relevant implications or at least merit discussion.
[1] As regards the interpretation of insurance exchange policies, see for example Asser/Wansink, Van Tiggele & Salomons 7-IX 2019, no. 347.
[2] As regards the VMZB 2021, see also the article by P.W.M. van Breukelen, published in October in the insurance exchange journal ‘Beursbengel’ and entitled ‘The new insurance exchange fire conditions: old wine in new bottles?’
[3] See https://www.vnab.nl/Pages/nl-NL/Algemeen/faq-nieuwe-modelvoorwaarden-brand. The presentation took place on 28 June 2021. Afterwards, the conditions were published on the VNAB website.
[4] Cf. the judgment of Rotterdam District Court of 21 January 2009, ECLI:NL:RBROT:2009:BH1979, in which it was held that there was no cover for corrosion as a cause of damage This meant that damage caused during a storm by the failure of a ship’s mast that had been weakened by corrosion was not covered. For the record, it should be noted that meither the NBUG nor the NBZB were applicable to the policy in question
[5] See the judgments of The Hague Court of Appeal of 31 January 2021, ECLI:NL:GHDHA:2021:111 and ECLI:NL:GHDHA:2021:110, on coverage under NBUG-inspired conditions for damage caused by a sinkhole/subsidence. For the sake of transparency, it should be noted that I myself was involved in those proceedings.
[6] See an answer to a question on the above-mentioned VNAB website to the effect that water accumulation is not included as standard cover.
[7] See the judgment of the Supreme Court of 9 June 2017, ECLI:NL:HR:2017:1055, legal ground 3.3.2 (Corrosieoven).
[8] As regards the named perils nature of the NBUG 1998, see, for example, the article by Drion, VA 1999, 2.
[9] See: ECLI:NL:HR:2021:815 in the case of Bosporus v ASR.
[10] Nor is this discussed in the case law published on this provision.
[11] Amsterdam Court of Appeal, 25 July 2017, ECLI:NL:GHAMS:2017:3852. This judgment was upheld by the Supreme Court in its judgment of 14 December 2017 (ECLI:NL:HR:2018:2261) for technical reasons concerned with the nature of appeals in cassation.
[12] See also the FAQ on the VNAB website about article 2.8 VMZB 2021: ‘For coverage under this article, the thing in question must first have functioned properly.’
[13] Cf. Supreme Court, 18 April 1986, ECLI:NL:HR:1986:AC9304, (ENCI v Lindelauf).
[14] See Amsterdam District Court 3 February 20210, ECLI:NL:RBAMS:2010:BM0448 on the question of whether a shipyard/gantry crane constitutes a ‘building’ for the purposes of the NBUG.