A better understanding of rock mass behavior is an essential part of the design and construction of underground structures. Any improvement in the understanding of the behavior of rock mass will facilitate the improvement of the design in terms of the safety of the working environment, long-term safety of the structure, environmental effects, and sound management of public or private resources. Thus, in step one in this paper the experience gained from applying the GDE (Geo Data Engineering) multiple graph approach for rock mass classification and assessing its behavior throughout excavation of the Alborz tunnel is presented. The predicted hazards are compared with the experienced problems and suggestions are given to be considered in future works of tunnel construction. In step two, the GDE multiple graph approach is compared to the rock mass behavior types proposed by Palmstrom & Stille (2007) in terms of the continuity of rock mass. The result of this comparison together with the data obtained from rock mass classification in the Alborz tunnel are used to develop a system that determines the applicability of the rock bolt supporting factor (RSF) in different rock mass behavior classes.

Budget overrun and schedule delay of infrastructure projects result in mismanagement of huge amounts of resources. Studies show that accounting for uncertainties in time and cost estimation of such projects can help the decision-makers in better understanding the involved risks. Several probabilistic models have been developed during the last two decades to facilitate such estimations. A common aspect of all these models is the subjective assessment of unit activities’ duration, which affects the accuracy of the estimated time and cost significantly. Despite this, the mechanism governing the variability of unit activities’ duration has seldom been studied. Thus, this paper focuses on addressing this gap by using a unique set of data from a tunnel project. The variability in unit activities’ duration (construction performance variability) is governed by three main components: typical performance variability, minor machinery delays, and minor performance delays. Using these components, a novel method for modeling construction performance variability in probabilistic time estimation of tunneling projects is proposed. The application of the proposed method is demonstrated through a case example with a discussion on its important aspects, advantages, and limitations. The findings of this paper offer a resource to improve the accuracy of time estimation for tunneling projects.

Time and cost estimation of tunnelingprojects is usually performed in a deterministic manner.However, because the deterministic approach isnot capable of dealing with uncertainty, probabilisticmethods have been developed over the years to betteraccount for this problem. Three models of this typeare the Decision Aids for Tunneling (DAT) and twomodels developed at KTH Royal Institute of Technologyand the Czech Technical University in Prague.To conduct a probabilistic time and cost estimation,it is important to understand and account for not onlythe uncertain factors that affect the project time andcost but also the involved parties’ different interestsand contractual responsibilities. This paper developsa risk model for the specific purpose of time andcost estimation of tunneling projects. In light of thismodel, the practical application of the three probabilisticmodels is discussed from a risk-aware decisionmaker’sperspective. The acquired insights can behelpful in increasing the experts’ risk-awareness inmodeling time and cost of tunneling projects.

Transport infrastructure projects frequently encounter challenges such as schedule delays and cost overruns, leading to substantial misallocation of public or private resources. These issues are often exacerbated by uncertainties in time and cost estimation outcomes. To address this, various models have been developed in recent years to enable probabilistic estimation for tunneling projects, considering uncertainties. However, the impact of uncertainties on the critical path and its implications on time and cost estimations have not been discussed explicitly and in detail for network underground structures encompassing multiple construction paths. In this study, we update the KTH time and cost estimation model using the Markov Chain Monte Carlo (MCMC) method. This updated model enables simultaneous round-by-round construction simulation across all paths in network underground structures. Consequently, it accommodates uncertainty in the critical path and its influence on time estimation outcomes. Additionally, the updated model introduces an innovative technique to model geological uncertainties along tunnel routes, thereby contributing to the field's diversity. Practical application of the updated model is showcased using the Uri Hearace tunnel as an illustrative example. The paper also delves into the practical implications of the findings from a decision-maker's standpoint, as well as discussing the model's limitations.

Cost overruns and schedule delays are frequently observed occurrences in the construction of transport infrastructure projects. Such phenomena lead to the mismanagement of significant amounts of both public and private resources.An examination of the literature reveals that uncertainty stands out as one of the potential primary causes of cost overruns and schedule delays. To address the impact of uncertainty on time and cost estimations in transport infrastructure projects, probabilistic approaches can be employed. 

In this doctoral thesis, first a conceptual risk model has been formulated specifically for the purpose of enhancing time and cost estimations in tunneling projects. This risk model serves as a tool to scrutinize and contrast existing probabilistic time and cost estimation models for tunnel projects, aiming to identify potential areas for improvement. Furthermore, the conceptual model is utilized to delve into the factors influencing the accuracy of subjective assessments regarding the input parameters in time estimation models. It also explores methods for incorporating the role of tunneling phases into the subjective assessment of these input parameters.

Then, enhancements and updates are introduced to the existingKTH model for time and cost estimation in tunneling projects. This model primarily targets three main sources of uncertainty: variability in construction performance, geological uncertainties, and the potential incidence of disruptive events. The analysis and improvements related to modelling of construction performance involve three sequential steps. In the first step, the construction process is modeled using the work breakdown structure (WBS), enabling a more realistic assessment of tunneling time. Subsequently, in the second step, PERT distributions are employed to model the uncertainty in the duration of unit activities, compared to the commonly used triangular distributions. The third step involves a detailed examination of a real tunnelling project's data to identify components contributing to construction performance variability for unit activities. This analysis pinpoints three main components: typical performance variability, minor performance delays, and minor machinery delays. These components are integrated into the KTH model, resulting in its further update concerning construction performance variability. 

A novel approach is introduced into the KTH model by leveraging the Metropolis-Hastings (MH) algorithm within the framework of Markov Chain Monte Carlo (MCMC) simulation to address geological uncertainties along the tunnel route. This method facilitates round-by-round simulation of the tunneling process and allows the model to accommodate uncertainty in the critical path for tunneling projects involving multiple headings. These enhancements aim to improve decision-making processes and mitigate risks associated with schedule delays and cost overruns. Additionally, the magnitude of disruptive events are now modeled as stochastic variables, an improvement on the original version of the KTH model.

Kostnadsöverskridanden och förseningar i tidplanen inträffar ofta vidbyggande av transportinfrastrukturprojekt. Detta leder till slöseri av betydande resurser, både offentliga och privata. En genomgångav litteraturen visar att osäkerhet framträder som en av de potentiella primära orsakerna. För att hantera påverkan av osäkerhet på tid- och kostnadsuppskattningar i transportinfrastrukturprojekt kan probabilistiska metoder användas. I denna doktorsavhandling utarbetades först en konceptuell riskmodell som kan förbättra tid- och kostnadsuppskattningar specifikt i tunnelprojekt. Riskmodellen användes som ett verktyg för att granska och jämföra olika befintliga probabilistiska modeller för sådana skattningar, i syfte att identifiera möjliga förbättringsområden. Riskmodellen användes också för att undersöka faktorer som påverkar noggrannheten i subjektiva bedömningar av indata i tidsuppskattningsmodeller. Även metoder för att inkludera olika tunnelbyggnadsfaser i den subjektiva bedömningen av dessa indata utforskades. Forskningen har resulterat i förbättringar och uppdateringar av den befintliga KTH modellen för tid- och kostnadsuppskattning i tunnelprojekt. Modellen inriktar sig mot tre huvudsakliga källor till osäkerhet i skattningen: variabilitet i arbetsprestation, geologisk osäkerhet och förekomstav försenande händelser. Analysen och förbättringarna av modelleringen av arbetsprestation utfördes i tre steg. I det första stegetmodellerades byggprocessen med hjälp av en så kallad Work Breakdown Structure (WBS), vilket möjliggör en mer realistisk bedömningav tunnelprojektets byggtid. I det andra steget användes PERTfördelningar för att modellera osäkerheten i tidsåtgång för de olika aktiviteterna i produktionscykeln, istället för den annars ofta använda triangelfördelningen. Det tredje steget utgjordes av en detaljerad undersökning av data från ett verkligt tunnelprojekt föratt identifiera vilka komponenter som bidrar till variabiliteten i arbetsprestation i produktionscykelns olika aktiviteter. Denna analyspekar ut tre huvudkomponenter: typisk variabilitet i arbetsprestation, mindre prestationsförseningar och mindre maskinförseningar. Dessa komponenter integrerades i KTH-modellen, vilket resulteradei ytterligare uppdateringar avseende variabiliteten i arbetsprestation. En ny metod infördes i KTH-modellen genom att använda Metropolis-Hastings-algoritmen inom ramen för Monte Carlo-simuleringmed Markovkedjor, för att hantera geologiska osäkerheter längstunnelsträckningen. Denna metod möjliggör stegvis simulering av tunnelbyggnadsprocessen så att KTH-modellen nu kan beakta osäkerhet i kritiska linjen i tunnelprojekt med flera fronter. Dessa förbättringar syftar till att underlätta beslutsfattandet och minska riskerna för förseningar och kostnadsöverskridanden. Dessutom är det nu möjligt att modellera storleken på försenande händelser som stokastiska variabler, vilket är en annan förbättring jämfört med den ursprungliga versionen av KTH-modellen. 

The PERT distribution may be a suitable distribution for modeling activities’ duration in probabilistic time estimation of tunnel projects as it puts more emphasis on the mean value of the distribution. In this paper, we compared the outcome of time estimations for a tunnel, using the triangular and PERT distributions for modeling the uncertainty of activities’ duration. The results indicate that the choice of the distribution affects the total estimated time considerably. In addition, the skewness of the distribution also affects the results of estimation meaning that realistic assessment of the parameters of the distributions is important.

Probabilistic time estimation is an essential part of proper risk management in tunneling projects. In recent decades, several models have been developed for this purpose, one of which was developed by Isaksson and Stille (Rock Mech Rock Eng 38:373–398, 2005). In this paper, Isaksson and Stille’s probabilistic time and cost estimation model was improved and then applied to estimate the total tunneling time of the headrace tunnel in the Uri hydropower project in India. The improvements allow the user to more accurately account for different types of geological features and disruptive events. The result of the estimation is a distribution of tunneling time. The outcome illustrates how a proper understanding of the geological setting of the project and its effect on construction performance can contribute to effective risk management. 

Rock Mass Rating (RMR) system has been applied in numerous civil and mining engineering projects. However, the applicability of the system in underground intersections has not yet been discussed. By creating an underground intersection, the only parameter of RMR that is affected is joint orientation favorability as the other five parameters (i.e. uniaxial compressive strength of rock material, joint spacing, RQD, groundwater condition and condition of joints) are not structural features but the intrinsic properties of rock mass. The suggested approach is based on considering the minimum rating for joint orientation favorability in the intersections as any set of discontinuity would be parallel or close to parallel with one of the openings in intersection. The application process and usefulness of suggested method is demonstrated by using it in selection of the most suitable location among 5 alternative options for starting a middle excavation face through an adit in the Alborz Tunnel Project. The case example shows that the method can be very helpful and important in practical applications of RMR system. 

I inledningen av varje tunnelprojekt behövs en bedömning av projekttid och projektets kostnad, som är korrekt och väl avvägd. Vad vi ser i dag, både i Sverige och internationellt, är dock att en överskriden budget och förseningar snarare är regel än undantag. Vi på KTH Jord- och bergmekanik har därför initierat en serie forskningsprojekt med finansiering av Formas och BeFo. Syftet är att förbättra befintliga metoder för dessa bedömningar, samt stärka möjligheterna till en rättvis riskdelning mellan parterna. Denna artikel ger en översikt över de idag existerande sannolikhetsbaserade tid- och kostnadsbedömningsmetoderna för tunnelbyggnad.

Transport infrastructure projects, including tunneling, suffer from timedelays and cost overrun. A literature review shows that the effect ofuncertainty has been neglected in explaining time and cost overrunmeaning that technical explanations matter. Probabilistic estimations oftime and cost can be employed for dealing with uncertainty in transportinfrastructure projects.In this licentiate thesis, KTH’s probabilistic time and cost estimationmodel for tunneling projects (Isaksson and Stille, 2005, Rock Mech. RockEng., 38, 373-398) was improved. The improvements include breakingdown the production activities into their sub-activities, which form thebasis for assessing times (or costs) for tunnel construction. In addition, theexceptional time and the length of model’s geotechnical zones aredescribed as stochastic variables instead of deterministic values used in theoriginal model. Given its hierarchical structure, the model can be used fortime and cost estimation of all types of tunnels and all constructionmethods in various geological condition.The improved version of the model uses three types of input parametersthat are probabilities of occurrences of different geological condition andidentified undesirable events, production effort of sub-activities (i.e. timespent for performing the sub-activity per unit length of tunnel) andadditive delay time that is imposed as a result of occurrence of undesirableevents. The important issues in modeling the uncertainty in the productionefforts of sub-activities are also explained.

Transportinfrastrukturprojekt, inklusive tunnelbyggnad, lider ofta avförseningar och ökade kostnader. En litteraturgenomgång visar atteffekten av osäkerhet inte har beaktats när man försöker förklara orsakentill förseningar och kostnadsökningar, vilket betyder att osäkerheten omgeotekniska förhållanden mycket väl kan spela en stor roll. För att hanteradenna osäkerhet när man bedömer tid och kostnad för tunnelprojekt kansannolikhetsbaserade metoder användas.I denna licentiatuppsats förbättrades den sannolikhetsbaserade modellför tid och kostnadsskattning i tunnelprojekt som tidigare utvecklats påKTH (Isaksson och Stille, 2005, Rock Mech. Rock Eng., 38, 373-398). Denviktigaste förbättringen var att dela upp produktionsaktiviteterna idelaktiviteter, för vilka man sedan enklare kan bedöma tidsåtgång (ellerkostnader). Dessutom beskrivs exceptionella förseningar och längden påmodellens geotekniska zoner nu med stokastiska variabler istället för meddeterministiska värden. Modellen är flexibel nog att kunna användas förtids- och kostnadsskattning av alla typer av tunnelprojekt ochkonstruktionsmetoder i olika geologiska miljöer.Den förbättrade versionen av modellen använder tre typer avindataparametrar: sannolikheter för förekomster av olika geologiskatillstånd och identifierade oönskade händelser; produktionsinsats fördelaktiviteter (d.v.s. den tid som används för att utföra delaktiviteten permeter tunnel); samt försening som orsakas av oönskade händelser.Uppsatsen diskuterar även de viktigaste aspekterna vid modellering avosäkerheten i produktionsinsatsen för delaktiviteter.