集群式開發(fā)方案在大型油氣藏開發(fā)中將發(fā)揮重大作用！

編譯 | TOM 驚蟄

在多勘探區(qū)油氣藏中，如果開發(fā)商僅從經(jīng)濟(jì)學(xué)角度來證明每個勘探區(qū)開發(fā)的合理性，常常會使自己處于非常不利的位置。在集群式開發(fā)方法中，將不同探勘區(qū)之間的地質(zhì)相關(guān)性與勘探區(qū)的集群相結(jié)合，以確定開發(fā)的經(jīng)濟(jì)可行性。雖然方法很直觀，但對集群式開發(fā)的經(jīng)濟(jì)評估通常不考慮實際存在的地質(zhì)相關(guān)性以及開發(fā)的協(xié)同效應(yīng)。對這些協(xié)同效應(yīng)進(jìn)行建模分析，發(fā)現(xiàn)它們對油氣藏開發(fā)的經(jīng)濟(jì)評估具有顯著影響。

在評估不同勘探區(qū)之間相關(guān)性的經(jīng)濟(jì)影響時，必須將地質(zhì)勘探成果分解為共享要素與原位要素。所有相關(guān)勘探區(qū)之間共享的地質(zhì)要素稱為共享或油氣藏機(jī)會。在每個相關(guān)勘探區(qū)之間獨(dú)立地質(zhì)機(jī)會要素被稱為前景/成功率（PSR）。任何個體勘探區(qū)的地質(zhì)成功率將被定義為油氣藏機(jī)會乘以PSR。

在評估新區(qū)塊時，若該區(qū)塊地質(zhì)機(jī)會要素相關(guān)性占比較高，將有效地使完全獨(dú)立的要素?zé)o法發(fā)揮應(yīng)有的作用。從研究人員的油氣藏分析中可獲得關(guān)鍵信息，油氣藏開發(fā)存在最優(yōu)鉆井順序。對油氣藏機(jī)會影響最大的井，通常是首先開鉆的最優(yōu)井。行業(yè)慣例一般是側(cè)重于井的決策，而不是油氣藏的決策。因此，常規(guī)做法是首先鉆探地質(zhì)成功機(jī)會（Pg）最大或平均可采原油最多的井。而理想情況下，應(yīng)該鉆探油氣藏機(jī)會要素最多的勘探區(qū)。

地質(zhì)上的成功并非都是商業(yè)或是經(jīng)濟(jì)上的成功。為確定油氣藏能否可以進(jìn)行商業(yè)化開發(fā)，需經(jīng)過兩步流程。第一步是評估Pg機(jī)會，第二步是運(yùn)行現(xiàn)金流模型，以確定每個勘探區(qū)具有商業(yè)價值的油田最小規(guī)模（MCFS），假設(shè)獨(dú)立開發(fā)以及現(xiàn)有開發(fā)的集群式綁定。兩者的產(chǎn)物即為商業(yè)成功機(jī)會。將油氣藏視為個體勘探區(qū)或集群勘探區(qū)，將在決定商業(yè)化規(guī)模的閾值時，對決策產(chǎn)生重大影響。假設(shè)每個勘探區(qū)都可以用相同的現(xiàn)金流模型(勘探區(qū)到中央處理設(shè)備的距離相同)。

需要認(rèn)識到非常規(guī)油氣藏經(jīng)濟(jì)評價的主要因素，即開發(fā)決策，是基于平均項目產(chǎn)量，而不是單口井的產(chǎn)量。在油氣藏中，眾數(shù)或單口井最可能的產(chǎn)量（從頻率角度來看）通常是不具經(jīng)濟(jì)效益的。目前，正在開采的油氣藏都每口井的平均產(chǎn)量是具有商業(yè)價值的。隨著累計井?dāng)?shù)的增加，方差（P10與P90比率）降低，進(jìn)而趨于正態(tài)分布。

隨著累計井?dāng)?shù)的增加，對數(shù)正態(tài)分布的眾數(shù)逐漸轉(zhuǎn)變?yōu)槠骄鶖?shù)。在對數(shù)正態(tài)分布中，平均值將始終超過中值，始終大于眾數(shù)。在正態(tài)分布中，眾數(shù)等于中值，等于平均值。同樣的道理可以應(yīng)用到集群式開發(fā)新油藏的建模中。

為了說明集群開發(fā)方法對經(jīng)濟(jì)評估與決策過程的影響，研究人員提出了深水環(huán)境中三個勘探區(qū)的案例研究。在實際情況中，勘探區(qū)可能會有不同的油氣儲量分布與原位地質(zhì)機(jī)會因素。然而，在案例研究中假設(shè)所有勘探區(qū)的上述兩個參數(shù)都相同。這種簡化的目的是將重點(diǎn)放在過程上，而不是放在單個勘探區(qū)的細(xì)節(jié)上。同樣，案例研究不是僅限于深水環(huán)境，該過程適用于任何環(huán)境，無論是陸上還是海上。

雖然現(xiàn)金流模型會影響經(jīng)濟(jì)結(jié)果，但它被認(rèn)為獨(dú)立于本文提出的集群式開發(fā)的評估流程，并沒有在評估中凸顯作用。目前提出的評估流程將與其他財政制度合作，例如產(chǎn)量分享合同模式。

業(yè)內(nèi)的傳統(tǒng)方法是，一個新區(qū)域的評估是基于足以支持基礎(chǔ)設(shè)施建設(shè)的大發(fā)現(xiàn)。在不考慮地質(zhì)相關(guān)性影響的集群式開發(fā)的基礎(chǔ)上，做出開發(fā)決策的分析。集群式開發(fā)擁有地質(zhì)相關(guān)性的勘探區(qū)。然后，綜合考慮所有三個勘探區(qū)中可能的發(fā)現(xiàn)，制定開發(fā)決策。研究人員提倡第三種方法，下面將對此進(jìn)行討論。

具有共享油氣藏機(jī)會的集群式捆綁開發(fā)。研究人員認(rèn)為，深水案例是一個經(jīng)過驗證的斷塊油氣藏，共享油氣藏機(jī)會占比高達(dá)80%。在經(jīng)濟(jì)性分析中，假定干井分析不能確定共享油氣藏機(jī)會系數(shù)，并對三個勘探區(qū)進(jìn)行鉆井。這種保守的案例研究方法是有意為之，因為它仍可證明，考慮共享油氣藏機(jī)會能夠為經(jīng)濟(jì)分析增加價值。

對于擁有共享油氣藏機(jī)會的集群式捆綁開發(fā)方法，經(jīng)濟(jì)評估就類似于假定地質(zhì)獨(dú)立的勘探區(qū)所采用的方法。采用60000次迭代的蒙特卡洛計算法，同時考慮了勘探區(qū)之間的共享油氣藏機(jī)會，利用三種勘探區(qū)的地質(zhì)資源儲量分布模擬出可能的產(chǎn)量。

這需要一個九分支決策樹，而不是三個評估模型中用于第二個的八分支決策樹，來總結(jié)出模擬結(jié)果，因為會有兩種不同的方式導(dǎo)致徹底失?。河蜌獠夭淮嬖冢蛘呒词褂蜌獠卮嬖?，所有干井都是由原位勘探區(qū)地質(zhì)特征因素所引起的。

決策樹的主要結(jié)果顯示，多重發(fā)展的可能性現(xiàn)在為22.4％（與不考慮共享油氣藏機(jī)會時的19％相比）。此外，決策樹還顯示無法開發(fā)的可能性為63.3％（在不考慮共享油氣藏機(jī)會時略高于62.9％）。在對勘探區(qū)之間的地質(zhì)相關(guān)性進(jìn)行建模時，這些結(jié)果與預(yù)期結(jié)果一致。在一個勘探區(qū)中取得成功的好消息，將增加在另一個相關(guān)勘探區(qū)中取得成功的機(jī)會，從而增加多個開發(fā)的機(jī)會(并增加成功案例的油氣儲量)。然而，導(dǎo)致一個勘探區(qū)出現(xiàn)干井的壞消息，增加了其他相關(guān)勘探區(qū)出現(xiàn)干井的可能性，從而增加了該項目失敗的總機(jī)會。

由于具有開發(fā)價值的獨(dú)立勘探區(qū)規(guī)模的門檻較低，因此使用集群式開發(fā)，至少一個開發(fā)項目的概率已經(jīng)增加。在集群式開發(fā)的戰(zhàn)略下，多個開發(fā)項目的機(jī)會顯著增加?？紤]到共享油氣藏機(jī)會增加了成功案例中的開發(fā)數(shù)量，但由于地理上的相關(guān)性，會降低商業(yè)成功的機(jī)會。

當(dāng)放棄獨(dú)立策略時，平均開發(fā)數(shù)量、平均原油最終采收率以及平均DNCF設(shè)為8%，預(yù)期價值與成功案例指標(biāo)均有顯著改善。當(dāng)考慮到成功案例中開發(fā)的更多發(fā)現(xiàn)所帶來的共享油氣藏機(jī)會時，集群式開發(fā)方法也顯示出經(jīng)濟(jì)上的改善，而整體開發(fā)機(jī)會只減少了很小一部分。

In plays with multiple prospects, decision-makers put themselves at a significant disadvantage by using economics on a standalone basis to justify development for each prospect. In the clustered-development approach, the geologic dependence between prospects is combined with the aggregation of prospects to determine the economic viability of a grassroots development offshore. Although easily visualized, economic evaluations of clustered developments typically do not consider the geological dependencies and development synergies that exist. Modeling these synergies has a significant effect on the economic evaluation of developments.

Geologic Dependency

In assessing the economic effect of the dependencies between prospects, breaking the chance of geologic success into a shared element and a local element is mandatory. The geologic-chance elements shared between all related prospects will be referred to as either the shared or play chance. Those geologic-chance elements that are independent between each related prospect are referred to as the prospect/success ratio (PSR). The probability of geologic success for any individual prospect will be defined as the play chance multiplied by the PSR.

When evaluating a new region, having a high degree of dependence within the geologic chance-of-success components will effectively derisk the play relative to total independence. A key learning from the authors’ play analysis is that an optimal drill order always exists. The well that will have the largest effect on the play chance is typically the optimal well to drill first. Industry practice often is focused on a well decision rather than a play decision. Consequently, the well with the highest chance of geologic success (Pg) or the largest risked mean recoverable resources typically is drilled first. Ideally, prospects should be drilled in a location that best addresses the play-chance elements.

Effect of Aggregation on Size Dependency

Geological successes are not all commercial or economic successes. A two-step process is used to determine the chance of commercial success leading to a development. The first step is assessing the total chance of Pg. The second step is to run the cash-flow model to determine the minimum commercial field size (MCFS) for each prospect, assuming a standalone development and a clustered tie-in to an existing development. The product of the two is the chance of commercial success. Viewing a play as an individual prospect vs. as an aggregate of prospects will affect decision making materially in the determination of the size threshold for commercialization. An assumption has been made that each prospect can be represented by the same cash-flow model (i.e., the same distance from prospect to central processing facilities).

One of the major factors in economic assessment of unconventional plays has been the understanding that development decisions are based on the average program outcome, not on individual well outcomes. In resource plays, the mode or most-likely single-well outcome (from a frequency perspective) typically is uneconomical. Resource plays that are active today are those in which the program’s average per well mean outcome is economical. As the number of wells being aggregated increases, the variance (P10/P90 ratio) decreases and the distribution tends toward a normal distribution.

The mode of the log-normal shifts toward the mean as the number of wells being aggregated increases. In a log--normal distribution, the mean will always exceed the median, which will always exceed the mode. In a normal distribution, the mode equals the median, which equals the mean. This same understanding can be applied in modeling clustered developments in greenfield plays.

Case Study

To illustrate the effect of the clustered--development approach on the economic-evaluation and decision-making process, a case study of three prospects in a deepwater setting is presented in the complete paper. In practice, the prospects likely would have different resource--volume distributions and local geologic-chance factors. For the case study in this paper, however, both parameters are assumed to be the same for all prospects. This simplification is meant to keep the focus on the process rather than on the details of the individual prospects. Likewise, the case study is not defined beyond being in a deepwater setting because the process is applicable in any setting, onshore or offshore.

The cash-flow model used to generate the economics assumes a royalty and tax fiscal regime. While the cash-flow model will affect the economic results, it is considered to be independent of the process of evaluating clustered developments being presented in this paper and is not emphasized. The process being presented will work with other fiscal regimes, such as a production-sharing-contract model.

The traditional industry approach, in which a new region is evaluated on the basis of one discovery large enough to support the infrastructure development.

An analysis wherein the development decision is made on the basis of a clustered development with no consideration for the effect of geologic dependence.

A clustered development with geologic dependence between the prospects. The development decisions are then based on what might be found when all three prospects are considered. The authors advocate this third approach, and it is discussed next.

Clustered Tie-In Development Methodology With Shared Play Chance. The authors consider that the deepwater case study is in a play-segment extension of a proven play with a high shared play chance of 80% associated with migration risk. For economic analysis, the assumption is made that dry-hole analysis will be inconclusive of the shared play chance factor and all three prospects will be drilled. This conservative approach for the case study is adopted deliberately because it still will demonstrate that consideration for the shared play chance will add value to the economic analysis.

For the clustered tie-in development methodology with shared play chance, economic evaluation is similar to that adopted when prospects are assumed to be geologically independent. A Monte Carlo simulation with 60,000 iterations was used to simulate the possible outcomes using the prospect geologic-resources distributions of the three prospects but now also considering the shared play chance between prospects.

A nine-branch decision tree is needed this time, rather than the eight-branch decision tree used in the second of the three evaluation models, to summarize the simulation results because total failure can now happen in two separate ways: the play does not exist or all dry holes are caused by local prospect-specific factors even though the play exists.

The main results of the decision tree show that the chance of multiple developments is now 22.4% (compared with 19% when not considering the shared play chance). The decision tree also shows that the chance of no developments is 63.3% (slightly higher than 62.9% when not considering the shared play chance). These results are consistent with expected outcomes when modeling geologic dependencies between prospects. Good news leading to success in one prospect will increase the chance of success in the other dependent prospects, increasing the chance of multiple developments (and increasing the success-case resource volumes). However, bad news leading to a dry hole in one prospect increases the chance of a dry hole in the other dependent prospects, increasing the overall chance of failure for the program.

The probability of having at least one development has increased with the use of clustered development because of the lower threshold on individual prospect discovery sizes that can be developed. The chance of multiple developments has increased significantly with a clustered-development strategy. Considering the shared play chance increases the number of developments in the success case but reduces the chance of commercial success because of the geologic dependencies.

Significant improvement in expected-value and success-case metrics is seen for the mean number of developments, mean oil estimated ultimate recovery, and mean DNCF at 8% when the standalone strategy is abandoned. The clustered-development approach also shows economic improvement when considering shared play chance caused by a larger number of discoveries developed in the success cases with only a small reduction in the overall chance of development.

備注：數(shù)據(jù)僅供參考，不作為投資依據(jù)。