CDP. Module: Introduction. Page: W0. Introduction. Water 2017 Information Request W0.1. Introduction

Transcription

1 CDP Water 2017 Information Request Microsoft Corporation Module: Introduction Page: W0. Introduction W0.1 Introduction Please give a general description and introduction to your organization At Microsoft, our mission is to empower every person and every organization on the planet to achieve more. Our strategy is to build best-in-class platforms and productivity services for a mobile-first, cloud-first world. We aim to reinvent productivity and business processes, build the intelligent cloud platform, and create more personal computing. Our unwavering commitment to responsible business practices and strong corporate governance is mission based and values driven. We know that water is a serious challenge that requires a global response from all sectors of society. All around the world, water resources are under increasing stress due to the combined effects of three significant challenges: population growth, economic development, and climate change. These trends, in turn, are driving an imbalance between water supply and demand that could jeopardize human health, agricultural productivity, economic development, and the ability to maintain sustainable ecosystems. In its 2017 Global Risks Report, the World Economic Forum ranks potential impacts from water crises higher than any other risk, including climate change, fiscal crises, and infectious disease. In FY16 (the reporting period for this response), we developed a new water stewardship strategy that builds on our belief that we should be looking at water in a more holistic way. This strategy will help us understand and articulate the true value of water in our operations, as well as the relative risks as water becomes scarcer. It focuses on increasing our understanding of water-related risks and impacts to our business and to the communities we serve, improving our water efficiency across datacenter, real estate, and manufacturing locations, and using our learnings to advance innovative solutions to water challenges. Our water strategy operates in parallel with our ongoing commitments to reducing our carbon footprint and searching for opportunities to minimize our environmental impact, reduce waste, and conserve other raw materials. In pursuing these goals, we follow strict policies to ensure that the company remains fully compliant with international environmental regulations and the specific environmental requirements of each country/region where we do business. W0.2 Reporting year Please state the start and end date of the year for which you are reporting data

2 Period for which data is reported Wed 01 Jul Thu 30 Jun 2016 W0.3 Reporting boundary Please indicate the category that describes the reporting boundary for companies, entities, or groups for which water-related impacts are reported Companies, entities or groups over which operational control is exercised W0.4 Exclusions Are there any geographies, facilities or types of water inputs/outputs within this boundary which are not included in your disclosure? No W0.4a Exclusions Please report the exclusions in the following table Exclusion Please explain why you have made the exclusion

3 Further Information Note that information provided on the Microsoft manufacturing supply chain reflects a different reporting period than for Microsoft direct operations. Our manufacturing supply chain information is based on an annual survey that is conducted on a calendar year basis, rather than the fiscal year reporting used for direct operations. For this response, the supply chain information reflects calendar year 2015, because the 2016 data was not yet available at the time of response submission. In addition, we collect monthly water reporting data from our Tier 1 manufacturing suppliers. Module: Current State Page: W1. Context W1.1 Please rate the importance (current and future) of water quality and water quantity to the success of your organization Water quality and quantity Direct use importance rating Indirect use importance rating Please explain Sufficient amounts of good quality freshwater available for use Sufficient amounts of recycled, brackish and/or produced water available for use Vital for operations Important Important Important Depending on the cooling technology deployed, for some datacenters, access to makeup water for cooling systems is vital for the continuous delivery of customer services. Where possible, we prefer to rely on free cooling using outside air when temperatures/humidity allow. This is often possible during months when temperatures are 75 Fahrenheit. Where we must rely on freshwater for cooling, we use on-site storage or multiple connections to the utility to help ensure that there is no disruption to the availability of this water. For other facilities (e.g. offices, labs), access to freshwater is required for employee consumption, restrooms, and cooking (some locations). Our manufacturing sites and Tier 1 manufacturing suppliers do not use significant amounts of water. For nonmanufacturing/indirect suppliers, availability of freshwater (quantity/quality) is not very important to our operations. Some datacenters use recycled water as an alternative makeup water source for cooling systems. Our preference is to rely on free cooling using outside air when temperatures/humidity allow, such as when temperatures are 75 F. When process water is required, we take advantage of recycled, reused, or industrial water where available unless we require domestic supply (e.g. with adiabatic

4 Water quality and quantity Direct use importance rating Indirect use importance rating Please explain cooling systems). Therefore, the direct use of recycled water is important yet only occasionally used in our operations. For our other facilities (e.g. offices, labs, manufacturing sites), these types of water are not used other than for nonessential landscaping. Almost half our manufacturing suppliers report using recycled or reused water as at least one source of water withdrawal; none of our manufacturing suppliers report using brackish water. For nonmanufacturing/indirect suppliers, availability of recycled, brackish, and/or produced water (quantity/quality) is not very important to our operations. W1.2 For your total operations, please detail which of the following water aspects are regularly measured and monitored and provide an explanation as to why or why not Water aspect % of sites/facilities/operations Please explain Water withdrawals- total volumes Water withdrawalsvolume by sources Water discharges- total volumes We are reporting water withdrawals for 100 percent of our facilities. Water withdrawals are based on data from utility bills from our largest sites and, in some cases, estimations. We have a robust estimation methodology for leased sites that takes into account square footage, electricity consumption, and building type. We are reporting water withdrawals for 100 percent of our facilities. Water withdrawals are based on data from utility bills from our largest sites and, in some cases, estimations. We have a robust estimation methodology for leased sites that takes into account square footage, electricity consumption, and building type. The vast majority of our metered withdrawals come from municipal sources. Where water withdrawals are estimated, we assume they come from municipal sources. Most of our sites do not have discharge meters. In most cases, water consumption is low, and so for these sites we know that discharges are close to withdrawals. Where there is water consumption (such as for landscaping, evaporative coolers, cooling towers, settling ponds), we adjust discharges accordingly.

5 Water aspect % of sites/facilities/operations Please explain Water discharges- volume by destination Water discharges- volume by treatment method Water discharge quality data- quality by standard effluent parameters Water consumption- total volume Facilities providing fullyfunctioning WASH services for all workers Most of our sites do not have discharge meters. In most cases, water consumption is low, and so for these sites we know that discharges are close to withdrawals. Where there is water consumption (such as for landscaping, evaporative coolers, cooling towers, settling ponds), we adjust discharges accordingly. Most of our sites do not have discharge meters. In most cases, water consumption is low, and so for these sites we know that discharges are close to withdrawals. Where there is water consumption (such as for landscaping, evaporative coolers, cooling towers, settling ponds), we adjust discharges accordingly. Water discharge quality is only measured at some of our facilities, as the majority of our discharges are directed to municipal treatment plants. As part of this reporting exercise, we are identifying gaps in reporting capabilities and will be performing an analysis on what it would cost to add infrastructure to have the ability to report on this in the future. Most of our sites are offices that do not have discharge meters. For these sites, water consumption is zero unless they have landscaping that requires irrigation. Some datacenters have evaporative cooling or landscape irrigation Microsoft provides fully functioning WASH services for all workers at all of our sites. W1.2a Water withdrawals: for the reporting year, please provide total water withdrawal data by source, across your operations

6 Source Quantity (megaliters/year) How does total water withdrawals for this source compare to the last reporting year? Comment Fresh surface water 0 Not applicable Microsoft does not withdraw any fresh surface water. Brackish surface water/seawater 0 Not applicable Microsoft does not withdraw any brackish surface water. Rainwater 0.94 About the same Despite expanding our water data collection coverage this reporting year, our total measured rainwater use was nearly the same as it was during the previous reporting period. Groundwater - renewable Groundwater - nonrenewable Produced/process water About the same Despite expanding our water data collection coverage this reporting year, our total measured renewable groundwater use was nearly the same as it was during the previous reporting period. 0 Not applicable Microsoft does not withdraw any nonrenewable groundwater. 0 Not applicable Microsoft does not use any produced/process water. Municipal supply About the same Wastewater from another organization About the same Total About the same In FY16 we began estimating water withdrawals for unmetered sites in order to measure the full scope of our water impact. For consistency and year-on-year tracking, we applied our estimation methodology to the previous year's water inventories. Once we applied the estimation methodology to our FY15 inventory (the previous reporting period), our total measured water withdrawals were nearly the same as for FY16 (the reporting period for this response). A large majority of water withdrawals comes from municipal supply. Despite expanding our water data collection coverage this reporting year, our total measured water withdrawals were nearly the same as they were during the previous reporting period. Municipally treated wastewater is used for landscape irrigation at some of our sites. In FY16 we began estimating water withdrawals for unmetered sites in order to measure the full scope of our water impact. For consistency and year-on-year tracking, we applied our estimation methodology to the previous year's water inventories. Once we applied the estimation methodology to our FY15 inventory (the previous reporting period), our total measured water withdrawals were nearly the same as for FY16 (the reporting period for this response). A large majority of water withdrawals comes from municipal supply. W1.2b

7 Water discharges: for the reporting year, please provide total water discharge data by destination, across your operations Destination Quantity (megaliters/year) How does total water discharged to this destination compare to the last reporting year? Comment Fresh surface water 0 Not applicable Microsoft does not discharge any water to fresh surface water sources. Brackish surface water/seawater 0 Not applicable Microsoft does not discharge any water to brackish surface water sources. Groundwater 0 Not applicable Microsoft does not discharge any water to groundwater sources. Municipal/industrial wastewater treatment plant Wastewater for another organization About the same For the majority of sites, we assume that withdrawals from municipal sources equal discharges to municipal sources. In FY16 we began estimating water withdrawals for unmetered sites in order to measure the full scope of our water impact. For consistency and year-on-year tracking, we applied our estimation methodology to the previous year's water inventories. Once we applied the estimation methodology to our FY15 inventory (the previous reporting period), our total measured water withdrawals were nearly the same as for FY16 (the reporting period for this response). 0 Not applicable Microsoft does not discharge any wastewater to be reclaimed by other organizations. Total About the same For the majority of sites, we assume that withdrawals from municipal sources equal discharges to municipal sources. In FY16 we began estimating water withdrawals for unmetered sites in order to measure the full scope of our water impact. For consistency and year-on-year tracking, we applied our estimation methodology to the previous year's water inventories. Once we applied the estimation methodology to our FY15 inventory (the previous reporting period), our total measured water withdrawals were nearly the same as for FY16 (the reporting period for this response). W1.2c Water consumption: for the reporting year, please provide total water consumption data, across your operations

8 Consumption (megaliters/year) How does this consumption figure compare to the last reporting year? Comment About the same Despite expanding our water data collection coverage this reporting year, our total measured water consumption was nearly the same as it was during the previous reporting period. Water is consumed during landscape irrigation at some of our sites. Most datacenter water consumption is not yet incorporated into this response. W1.3 Do you request your suppliers to report on their water use, risks and/or management? Yes W1.3a Please provide the proportion of suppliers you request to report on their water use, risks and/or management and the proportion of your procurement spend this represents Proportion of suppliers % Total procurement spend % Rationale for this coverage Less than 1% Microsoft has nearly 50,000 manufacturing (direct), nonmanufacturing (indirect), and datacenter suppliers. We use the aggregated results of our annual manufacturing supplier survey plus risk indices to assess our risks into the future. We collect reports from our

9 Proportion of suppliers % Total procurement spend % Rationale for this coverage key manufacturing suppliers, prioritizing suppliers that make up the majority of our manufacturing spend, including Tier 1 and Tier 2 suppliers. Our manufacturing supplier questionnaire covers water-related topics such as withdrawal data and whether the supplier s operations significantly affect water sources. Although we request a response to our annual survey, responding is not a contractual requirement. In 2016, we requested 67 suppliers representing 95 percent of our manufacturing spend to respond to our survey. We also request all of our Tier 1 manufacturing suppliers (representing 7 percent of our spend) to provide water data monthly, and we work with these suppliers to reduce water consumption by performing a water balance equation and mitigating inefficiencies. For nonmanufacturing/indirect suppliers, we have determined there to be a low level of risk. In FY16 we ran a CDP Supply Chain pilot for water, with 16 of 17 requested nonmanufacturing suppliers responding; our average score was 3 grades higher than the global supplier average. W1.3b Please choose the option that best explains why you do not request your suppliers to report on their water use, risks and/or management Primary reason Please explain W1.4 Has your organization experienced any detrimental impacts related to water in the reporting year? No W1.4a

10 Please describe the detrimental impacts experienced by your organization related to water in the reporting year Country River basin Impact driver Impact Description of impact Length of impact Overall financial impact Response strategy Description of response strategy W1.4b Please choose the option below that best explains why you do not know if your organization experienced any detrimental impacts related to water in the reporting year and any plans you have to investigate this in the future Primary reason Future plans Further Information Module: Risk Assessment Page: W2. Procedures and Requirements W2.1 Does your organization undertake a water-related risk assessment? Water risks are assessed

11 W2.2 Please select the options that best describe your procedures with regard to assessing water risks Risk assessment procedure Coverage Scale Please explain Comprehensive company-wide risk assessment Direct operations and supply chain All facilities and some suppliers We focus water risk analyses on operations over which we have operational control. In FY16, we initiated a comprehensive company-wide assessment of vulnerabilities to the physical impacts of climate change (including sea-level rise, flooding from intense precipitation, drought frequency/intensity) for a subset of high-value facilities in locations likely subject to these impacts; once complete, the findings will be integrated into our enterprise water strategy. We use the WRI Aqueduct tool to assess our facilities (such as offices, labs, manufacturing sites, datacenters) to determine whether sites in at-risk river basins are critical. For our datacenters, all owned sites are vetted prior to acquisition to ensure adequate water and sewer quality/capacity exist. In some cases, local utilities provide a reused or recycled industrial water supply; in others, well water or domestic supply is available. Where future expansion plans are unknown, we obtain commitment letters from the utilities indicating that they can provide for future demand. We do not locate datacenters where utility constraints would prevent us from operating those facilities. We engage with select manufacturing suppliers on water usage and resources, prioritizing Tier 1 suppliers. For our nonmanufacturing/indirect suppliers, we have determined there to be a low level of risk; in FY16, we ran a CDP Supply Chain pilot for water, with 16 of 17 requested nonmanufacturing suppliers responding. W2.3 Please state how frequently you undertake water risk assessments, at what geographical scale and how far into the future you consider risks for each assessment

12 Frequency Geographic scale How far into the future are risks considered? Comment Annually River basin 3 to 6 years Sporadically not defined Sporadically not defined Facility Facility >6 years >6 years Annually Facility >6 years Annually Country >6 years Annually Country Up to 1 year At present, we conduct annual water risk assessments that consider the near future for facilities in our Real Estate & Facilities division (for example, offices and labs), datacenters, and manufacturing facilities using the WRI Aqueduct tool. From FY17, we also plan to use the Water Risk Monetizer (WRM) tool to quantify the financial risk associated with operating in water-scarce regions of the world (at the facility level). We initiated a comprehensive company-wide assessment of vulnerabilities to the physical impacts of climate change in FY16 (the reporting period). This assessment captures water-related impacts, including sea-level rise, flooding from intense precipitation events, drought frequency, and drought intensity. The hotspot analysis of potential impacts looks out to 2030 and 2060, where downscaled data is available. Prior to any land acquisition or lease, the datacenter engineering team evaluates the water and sewer capacity for that location. Water permits (if required) are obtained for both datacenters and manufacturing facilities through the local authority having jurisdiction (AHJ) or utility provider. We conduct an annual manufacturing supplier survey, because most of our manufacturing suppliers only have annual data/information available. We use the aggregated results of the survey plus risk indices to assess our risks into the future. We conduct an annual manufacturing supplier survey, because most of our manufacturing suppliers only have annual data/information available. We use the aggregated results of the survey plus risk indices to assess our risks into the future. For our nonmanufacturing suppliers, we assess water risk annually using the Verisk Maplecroft risk index in combination with spend, brand proximity, and 23 other indicators. W2.4 Have you evaluated how water risks could affect the success (viability, constraints) of your organization's growth strategy? Yes, evaluated over the next 10 years W2.4a Please explain how your organization evaluated the effects of water risks on the success (viability, constraints) of your organization's growth strategy?

13 We evaluate how water risks could affect the success of our growth strategy in different ways and for different time periods for different parts of our business. Overall, we are not a water-intensive company, and so water risks have not led to a change in our growth strategy. For example, our two most critical aspects of operations with respect to water are supplying employees with potable water and providing cooling for our datacenters. We believe potential mitigation efforts (bottled water, on-site filtration) can adequately mitigate risk in an event where potable water supply is interrupted. For our datacenters, in some cases, we have onsite storage capacity for supply should there be an interruption in water supply that could affect our ability to deliver continuous customer services (and therefore our growth strategy). Access to sufficient amounts of good quality freshwater, recycled, brackish, and/or produced water is important for some parts of our business, particularly for our manufacturing suppliers, and good quality freshwater is vital to our datacenter operations. In FY16, we initiated a comprehensive company-wide assessment of vulnerabilities to the physical impacts of climate change (including sea-level rise, flooding from intense precipitation, drought frequency/intensity) for a subset of high-value facilities in locations likely subject to these impacts; once complete, the findings will be integrated into our enterprise water management strategy. Each year we use the WRI Aqueduct tool to evaluate risk implications of water to our offices, labs, datacenters, and manufacturing sites as part of our annual CDP reporting. For our datacenters, we also evaluate the capacity of supply and discharge as part of our site selection due diligence, taking into account long-term (at least 10 years) operational requirements. We also take into consideration potential backup requirements. In FY16, we collected information related to water and our manufacturing suppliers operations for our Tier 1 suppliers. The results of this survey plus four risk indices help us evaluate risk implications of water to our manufacturing supply chain in the longer term (>6 years). For our nonmanufacturing suppliers, we assess water risk annually using the Verisk Maplecroft risk index in combination with spend, brand proximity and 23 other indicators. W2.4b What is the main reason for not having evaluated how water risks could affect the success (viability, constraints) of your organization's growth strategy, and are there any plans in place to do so in the future? Main reason Current plans Timeframe until evaluation Comment W2.5 Please state the methods used to assess water risks

14 Method Please explain how these methods are used in your risk assessment Internal company knowledge Maplecroft Global Water Security Risk Index WRI Aqueduct Other: ISO significant aspect and impacts review process; Environmental Performance Index; Corruptions Perceptions Index; Political Rights and Civil Liberties Score; Maplecroft; regional government databases; Federal Emergency Management Agency (FEMA) data; Intergovernmental Panel on Climate Change (IPCC) sea-level rise forecasts; proprietary modeled flood data We use the WRI Aqueduct tool to assess water risks for our Real Estate & Facilities division (for example, offices and labs), datacenters, and manufacturing sites because it reveals a broad spectrum of key water-related risks at the level of individual river basins. For our Real Estate & Facilities division, datacenters, and manufacturing sites, we import each of our locations into the tool by latitude/longitude to assess which sites are at risk and then vet results against internal company knowledge to determine which risks flagged by the tool are substantive to the business. To identify facility flood risk as part of our companywide vulnerability assessment to the physical impacts of climate change, we use FEMA flood data and IPCC data on sea-level rise and downscaled sea-level rise estimates provided by regional and local government authorities; we also engage with providers of proprietary flood modeling products. From FY17, we also plan to use the Water Risk Monetizer (WRM) tool to quantify the financial risk associated with operating in water-scarce regions of the world (at the facility level). Our datacenter engineering teams use internal company knowledge to evaluate supply/discharge capacity during site selection due diligence. The teams send requests for information (RFIs) containing water quality/quantity targets to economic development agencies in advance of touring potential sites. We ask the local water utility to confirm that they can meet those targets over time and tied to our phased deployment. Our engineers also review water quality reports to confirm the number of cycles of concentration that we can achieve, depending on the mechanical cooling solution type. We consider potential backup requirements and in some cases we have onsite storage capacity for supply should there be an interruption in utility service. Microsoft Manufacturing is certified to ISO and identifies substantive risk using the significant aspect and impacts review process; significance is determined by scoring the likelihood and magnitude of the impact on the environment, feasibility of addressing the impact, regulatory risk, resource scarcity, and community concerns. For our manufacturing suppliers, we use four risk indices (Maplecroft, Environmental Performance Index from Yale University, Corruptions Perceptions Index from Transparency International, and Political Rights and Civil Liberties Score from Freedom House) plus our annual supplier survey to assess water-related risks. For our nonmanufacturing/indirect suppliers, we assess water risk annually using the Verisk Maplecroft risk index in combination with spend, brand proximity, and 23 other indicators. W2.6

15 Which of the following contextual issues are always factored into your organization's water risk assessments? Issues Choose option Please explain Current water availability and quality parameters at a local level Current water regulatory frameworks and tariffs at a local level Current stakeholder conflicts concerning water resources at a local level Current implications of water on your key commodities/raw materials Current status of ecosystems and habitats at a local level Current river basin management plans Relevant, included Not relevant, explanation provided Not relevant, explanation provided Relevant, included Not relevant, included Not relevant, explanation provided When selecting locations, access to freshwater or industrial supply is a basic and fundamental criterion. For our existing locations, we use the WRI Aqueduct tool to assess the issue for our offices, labs, datacenters, and manufacturing sites. We use internal company knowledge (through site selection due diligence) to further assess the issue for our datacenters. From FY17, we also plan to use the Water Risk Monetizer (WRM) tool to quantify the financial risk associated with operating in water-scarce regions of the world (at the facility level). We do not yet conduct this analysis independently for our offices, labs, and manufacturing sites because they are not significant enough users of water to be exposed to unique regulations and tariffs compared with any other standard office building. For our datacenters, we explore this as part of site selection due diligence (method: internal company knowledge). Where permits or water rights are required, we obtain those as part of the process. The WRI Aqueduct tool currently does not assess the potential for stakeholder conflicts within water basins, and we do not yet conduct this analysis independently for our offices, labs, or manufacturing sites because they are not significant users of water. For our datacenters, water supply and discharge are preapproved during site selection due diligence processes (method: internal company knowledge); therefore, conflict resolution is not an issue at these sites because the risk has already been investigated. The effect of our manufacturing suppliers operations on water sources is covered in our annual supply chain survey. We use the aggregated results of the survey plus risk indices to assess our risks into the future. In addition, for key raw materials in our supply chain, we assess the current and future risk of impact on water sources at the materials extraction sites. For all other suppliers, the issue is not relevant because we do not source commodities/raw materials from these suppliers. Amphibian population health is a good indicator of the status of ecosystems. For our existing locations, we use the WRI Aqueduct tool (which assesses whether or not amphibians are threatened in watersheds) to assess this issue for our offices, labs, datacenters, and manufacturing sites. We do not see this as a relevant risk for our offices, labs, or manufacturing sites because they are not significant users of water. The WRI Aqueduct tool currently does not provide an indicator for river basin management plans, and we do not yet conduct this analysis independently for our offices, labs, or manufacturing sites. This issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years but is not undertaken now because these facilities do not use significant amounts of water. For our datacenters, we consult with local water providers and receive commitment letters from them indicating that they have the capacity required to meet future demand; in many locations, we rely on outside air cooling during the milder months and only draw water for cooling during the warmer months (daytime

16 Issues Choose option Please explain Current access to fully-functioning WASH services for all employees Estimates of future changes in water availability at a local level Estimates of future potential regulatory changes at a local level Estimates of future potential stakeholder conflicts at a local level Estimates of future implications of water on your key commodities/raw materials Relevant, included Relevant, included Not relevant, explanation provided Not relevant, explanation provided Relevant, included temperatures greater than 75 Fahrenheit). In locations where we do not use adiabatic cooling, the local water utility has sufficient capacity to meet seasonal demand such that it will not significantly impact the local water table in coordination with local agricultural needs. This risk is assessed through the current water availability metrics within the WRI Aqueduct tool. Facilities identified as being at risk for water shortages are also considered at risk for access to WASH services. For our existing locations, we use the WRI Aqueduct tool to assess this issue for our offices, labs, datacenters, and manufacturing sites. When selecting locations, ongoing access to freshwater is a basic and fundamental criterion. For our existing locations, we use the WRI Aqueduct tool to determine the basins in which we operate that are likely to experience water availability changes in the near future. For our datacenter projects, we engage with all utilities (water/sewer/power/fiber) prior to concluding a transaction. Capacity and quality are vetted and service agreements put in place prior to completing the transaction. Where future expansion plans are unknown, we obtain commitment letters from the utilities indicating that they can provide for future demand (often subject to expansion of existing infrastructure). We do not locate datacenters where utility constraints would prevent us from operating those facilities (method: internal company knowledge). In 2016, we began a comprehensive assessment of our vulnerabilities to the physical impacts of climate change, including the potential for diminished water availability at the local level; upon completion of this assessment we will integrate the results into our siting and operational planning as appropriate. From FY17, we also plan to use the Water Risk Monetizer (WRM) tool to quantify the financial risk associated with operating in water-scarce regions of the world (at the facility level). The WRI Aqueduct tool currently does not provide estimates of the potential for future regulatory changes within water basins, and we do not yet conduct this analysis independently. This issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years but is not undertaken now because our offices, labs, and manufacturing sites are not significant users of water. For our datacenters, regulatory changes are handled through permit revisions. The WRI Aqueduct tool currently does not assess the potential for future stakeholder conflicts within water basins, and we do not yet conduct this analysis independently for our offices, labs, or manufacturing sites because they are not significant users of water. For our datacenters, water supply and discharge are preapproved during site selection due diligence (method: internal company knowledge); we meet with key local representatives to determine the likelihood of future potential issues and whether a site is still viable or not. Therefore, conflict resolution is not an issue because the risk has already been investigated. The effect of our manufacturing suppliers operations on water sources is covered in our annual supply chain survey. We use the aggregated results of the survey plus risk indices to assess our risks into the future. In addition, for key raw materials in our supply chain, we assess the current and future risk of impact to water sources at the materials extraction sites. For all other suppliers, the issue is not relevant because we do not source commodities/raw materials from these suppliers.

17 Issues Choose option Please explain Estimates of future potential changes in the status of ecosystems and habitats at a local level Scenario analysis of availability of sufficient quantity and quality of water relevant for your operations at a local level Scenario analysis of regulatory and/or tariff changes at a local level Scenario analysis of stakeholder conflicts concerning water resources at a local level Scenario analysis of implications of water on your key commodities/raw materials Scenario analysis of potential changes in the status of ecosystems and habitats at a local level Not relevant, explanation provided Relevant, included Not relevant, explanation provided Not relevant, explanation provided Relevant, included Not relevant, explanation provided The WRI Aqueduct tool currently does not provide future estimates of the potential for changes in the status of ecosystems and habitats within water basins, and we do not yet conduct this analysis independently. This issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years but is not undertaken now because our offices, labs, and manufacturing sites are not significant users of water. For our datacenters, regulatory or ecosystem changes are handled through permit revisions. We conduct this analysis for our offices, labs, and manufacturing sites, where availability of sufficient quantity and quality of water is a prerequisite for any facility. For our datacenters, water supply and discharge are preapproved; therefore, quantity and quality requirements are vetted in advance. In 2016, we began a comprehensive assessment of our vulnerabilities to the physical impacts of climate change, including the potential for diminished water availability at the local level. The assessment includes a scenario analysis based on the IPCC RCP 8.5 scenario (mean 2 Celsius projected increase, relative to ). Upon completion of this assessment we will integrate the results into our siting and operational planning as appropriate. From FY17, we also plan to use the Water Risk Monetizer (WRM) tool to quantify the financial risk associated with operating in waterscarce regions of the world (at the facility level). The WRI Aqueduct tool currently does not provide estimates of the potential for future regulatory changes within water basins, and we do not yet conduct this analysis independently for our offices, labs, and manufacturing sites because these facilities are not significant users of water. For our datacenters, water supply and discharge are preapproved; therefore, quantity and quality requirements are vetted in advance, and tariffs and fees are included in any service agreement with the utility provider. Still, this issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years. The WRI Aqueduct tool currently does not assess the potential for future stakeholder conflicts within water basins, and we do not yet conduct this analysis independently for our offices, labs, and manufacturing sites because these facilities are not significant users of water. For our datacenters, water supply and discharge are preapproved; therefore, conflict resolution is not an issue. Still, this issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years. We conduct a scenario analysis if/when implications are reported for our manufacturing suppliers. For all other suppliers, the issue is not relevant because we do not source commodities/raw materials from these suppliers. The WRI Aqueduct tool currently does not provide future estimates of the potential for changes in the status of ecosystems and habitats within water basins, and we do not yet conduct this analysis independently for our offices, labs, and manufacturing sites because these facilities are not significant users of water. For our datacenters, regulatory or ecosystem changes are handled through permit revisions. Still, this issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years.

18 Issues Choose option Please explain Other W2.7 Which of the following stakeholders are always factored into your organization's water risk assessments? Stakeholder Choose option Please explain Customers Employees Investors Local communities NGOs Other water users at a local level Regulators Not relevant, explanation provided Relevant, included Relevant, included Not relevant, explanation provided Not relevant, explanation provided Relevant, included Relevant, included Customers are indirectly or directly relevant to everything Microsoft does. However, in this case, the customer impact of our operations is minimal and indirect, because our products are not water intensive in creation or use. Still, this issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years. When selecting locations, access to freshwater is a basic and fundamental criterion. Microsoft provides fully functioning WASH services for all workers at all of our sites. Access to freshwater is important for employee consumption, restrooms, and cooking (some locations). We have reported our water risks and responses publicly through CDP since 2012 at the request of investors. The health of the communities in which we operate is highly relevant for Microsoft. However, the water impact of our operations on local communities is minimal because we do not have water-intensive operations in the river basins in which we operate. Still, this issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years. The mission of NGOs seeking to preserve water resources is relevant to Microsoft. However, because we are not substantial water users in the river basins in which we operate, we have not directly considered, nor do we yet work with, NGOs in our water risk assessments. Still, this issue is under consideration for inclusion and will be evaluated in the next 6 to 10 years. For our offices, labs, and manufacturing sites, Microsoft is not a substantial water user in the river basins in which we operate. Therefore, we do not have a significant impact on the water sources for other water users. For datacenters, water supply and discharge are preapproved; therefore, quantity and quality requirements are vetted in advance and are often included in our service agreements with the local utility company. For our offices, labs, and manufacturing sites, Microsoft is not a substantial water user in the river basins in which we operate. Therefore, we do not have a significant interaction with water regulators in these locations. For datacenters, water supply and discharge are preapproved; therefore, quantity and quality requirements are vetted in advance and are often included in our service agreements with the local utility company.

19 Stakeholder Choose option Please explain River basin management authorities Statutory special interest groups at a local level Suppliers Water utilities at a local level Other Relevant, included Not relevant, explanation provided Relevant, included Relevant, included Only where water rights are required do we engage with bureau or basin management agencies. In those cases, the water permits are regulated and coordinated with those authorities. Microsoft is not a substantial water user in the river basins in which we operate. Therefore, we do not work with or otherwise consider statutory special interest groups in our water risk assessments. We monitor select manufacturing suppliers water management practices and wastewater compliance (prioritizing suppliers that make up the majority of our manufacturing spend, including Tier 1 and Tier 2 suppliers) to ensure that their water risks are minimized. Water is not a high risk for our nonmanufacturing/indirect suppliers and we do not monitor water disclosures for them. For our offices, labs, and manufacturing sites, Microsoft is not a substantial water user in the river basins in which we operate; therefore, since we have determined that there is no risk associated with these stakeholders, we do not consider water utilities any further in our water risk assessments for these facilities. For datacenters, water supply and discharge are preapproved; therefore, quantity and quality requirements are vetted in advance and are often included in our service agreements with the local utility company. W2.8 Please choose the option that best explains why your organisation does not undertake a water-related risk assessment Primary reason Please explain Further Information Module: Implications Page: W3. Water Risks

20 W3.1 Is your organization exposed to water risks, either current and/or future, that could generate a substantive change in your business, operations, revenue or expenditure? No W3.2 Please provide details as to how your organization defines substantive change in your business, operations, revenue or expenditure from water risk Risk assessments completed to date have determined that at present our business is not exposed to water risks that could generate a substantive change in our business, operations, revenue or expenditure. Substantive change for offices/labs would require significantly altering, closing or relocating the operations of a facility/group of facilities. This could include impacts on employees and/or technology equipment. Our operational analysis includes a 3-part filter to identify those facilities representing the greatest potential risk: 1) location in water basins with current or future potential for high baseline water stress; 2) annual average electricity use >60 kwh/square foot (large electricity users being datacenters); and 3) if the facility represents >1% of total companywide electricity use (using electricity as proxy for the size or relative importance of the facility to our company). We review our approach annually when we respond to the CDP water questionnaire, reassessing our risk assessment procedures and looking for opportunities to formalize/integrate risk assessments across the company. Similarly, substantive change for datacenters would entail the need to significantly alter, close or relocate a datacenter in response to water risk that would affect our ability to deliver continuous customer services. We engage all utilities (water/sewer/power/fiber) during site selection due diligence. Capacity and quality are vetted and service agreements put in place prior to completing the transaction. Where future expansion plans are unknown, we obtain commitment letters from the utilities indicating they can provide for future demand (often subject to expansion of existing infrastructure). We do not locate datacenters where utility constraints would prevent us from operating those facilities. Risk for manufacturing sites is defined per our ISO Significant Aspects and Impacts process and is not considered substantive. These definitions apply to direct operations only. In FY16, we initiated a comprehensive company-wide assessment of vulnerabilities to the physical impacts of climate change (including sea-level rise, flooding from intense precipitation, drought frequency/intensity), focusing on a subset of high-value facilities in locations likely subject to these impacts; once complete, the findings will be integrated into future risk assessments. W3.2a Please provide the number of facilities* per river basin exposed to water risks that could generate a substantive change in your business, operations, revenue or expenditure; and the proportion of company-widefacilities this represents

21 Country River basin Number of facilities exposed to water risk Proportion of companywide facilities that this represents (%) Comment W3.2b For each river basin mentioned in W3.2a, please provide the proportion of the company's total financial value that could be affected by water risks Country River basin Financial reporting metric Proportion of chosen metric that could be affected Comment W3.2c Please list the inherent water risks that could generate a substantive change in your business, operations, revenue or expenditure, the potential impact to your direct operations and the strategies to mitigate them Country River basin Risk driver Potential impact Description of potential impact Timeframe Likelihood Magnitude of potential financial impact Response strategy Costs of response strategy Details of strategy and costs

22 W3.2d Please list the inherent water risks that could generate a substantive change in your business operations, revenue or expenditure, the potential impact to your supply chain and the strategies to mitigate them Country River basin Risk driver Potential impact Description of potential impact Timeframe Likelihood Magnitude of potential financial impact Response strategy Costs of response strategy Details of strategy and costs W3.2e Please choose the option that best explains why you do not consider your organization to be exposed to water risks in your direct operations that could generate a substantive change in your business, operations, revenue or expenditure Primary reason Please explain Risks exist, but no substantive impact anticipated We are not a water-intensive business but are working to reduce water use. We use the WRI Aqueduct tool annually to identify office/lab, datacenter, and manufacturing locations with current/future potential for high baseline water stress and filter those results for our most important facilities. This analysis has not revealed any substantive water risk: each local operation is a relatively small contributor and most functions are mobile. Potential risks include water rationing, which would first impact landscaping and, in extreme cases, reduce work hours, though remote work locations would be available to employees. For our datacenters, capacity/quality are vetted with utilities and service agreements put in place prior to completing the transaction. Where future expansion plans are unknown, we obtain commitment letters from the utilities indicating that they can provide for future demand. We do not locate datacenters where utility constraints would prevent us from operating those facilities. The Quincy Bureau of Reclamation (Columbia River Basin) is the only US location where we use water rights for some of our datacenter water needs. Currently, the bureau has far more rights than users and we own more water rights than needed for cooling, so we have no substantive risk even here. We have built a reuse water treatment plant (using some of our infrastructure) and turned ownership over to the City of Quincy, which gives the city an abundant industrial water system.

23 W3.2f Please choose the option that best explains why you do not consider your organization to be exposed to water risks in your supply chain that could generate a substantive change in your business, operations, revenue or expenditure Primary reason Please explain Risks exist, but no substantive impact anticipated We conduct an annual manufacturing survey with suppliers that represent the majority of our spend and use the aggregated results of the survey plus four risk indices to assess our risks into the future. From this analysis, in combination with business importance and spend, we determined that there are no substantive water risks. For example, in 2015 (the most recent year for which we have data), some companies reported sharing their water source with other factories, which can create a risk for competition for resources; however, this is a nonsubstantive issue because there is ample municipal water supply and they are not located in a water-scarce area. This assessment will continue to take place annually. For our nonmanufacturing/indirect suppliers, we assess water risk annually using the Verisk Maplecroft risk index in combination with spend, brand proximity, and 23 other indicators. From this analysis we determined the risk to be low. W3.2g Please choose the option that best explains why you do not know if your organization is exposed to water risks that could generate a substantive change in your business operations, revenue or expenditure and discuss any future plans you have to assess this Primary reason Future plans Further Information Page: W4. Water Opportunities W4.1