Siemens Raftelis Phase 1 Report WM - Draft Final

1031 S. Caldwell Street Phone 704 . 373 . 1199 www.raftelis.
com
Suite 100 Fax 704 . 373 . 1113
Charlotte, NC 28203
February 27, 2015
Ms. Kishia Powell

Director, Department of Public Works
City of Jackson
200 South President Street
Jackson, MS 39201
Re: Phase 1 Final Report – Draft Final
om
Dear Ms. Powell:
.c
The Raftelis Financial Consultants, Inc. (RFC) Project Team (RFC, Intel Business Solutions, and SOL
ya
Engineering Services, LLC; and together, the Project Team) has completed our initial engagement for the
City of Jackson (City). The initial engagement evolved to include the following tasks.
la
 Task 1 – Strategic Financial Planning
ba
 Task 2 – Revenue Sufficiency Analysis

 Task 3 – Billing System Data Review
am
 Task 4 – Siemens Contract Review

 Task 5 – Infrastructure Master Plan Development
nj
 Task 6 – Additional Efforts

so
We have accomplished the objectives outlined in our scope of work and have assisted on a number of
ck
additional issues that have arisen during our engagement. The attached report, which consists of an
Executive Summary and four technical memoranda, summarizes our work and findings for the various
Ja
tasks.
We have appreciated the opportunity to work with you and your staff. As noted in the report, there are
a number of tasks that still need to be undertaken, and we hope to have an opportunity to continue to
work for the City and with you and your staff. Should you have any questions regarding the report,
please contact me at (704) 936-4433.
Sincerely,
Peiffer A. Brandt
Chief Operating Officer
EXECUTIVE SUMMARY
The City of Jackson (City) engaged a project team led by Raftelis Financial Consultants, Inc. (RFC) and
including Intel Business Solutions (IBS) and SOL Engineering (together, the RFC Project Team) to provide
assistance with various financial and management issues which the City currently faces. There were six
tasks included in the engagement. As the project progressed, the scope within certain tasks evolved
from what was originally contemplated. For example, the two tasks related to the Siemens contract
were combined into a single task. There was also an Additional Task, which included three subtasks
associated with issues that arose during the engagement, and was incorporated accordingly. A
summary of each task is provided below.
TASK 1 –STRATEGIC FINANCIAL PLANNING
om
The Project Team participated in two strategy sessions with City of Jackson leadership. Based on these
sessions, we identified 37 tasks the City needs to undertake to enhance the financial and management
sustainability of its Department of Public Works (DPW). A few of these tasks were included in the initial
.c
scope of work. We prioritized all the tasks, categorizing them as requiring attention 1) within six
ya
months; 2) in six months to a year; or 3) in a year plus. Of the 37 issues identified, 21 were categorized
as needing attention within six months. We prepared a brief technical memorandum (Technical
la
Memorandum 1 or TM1) summarizing our findings, which is part of this report. In addition, an Excel file
ba
with the matrix of prioritized tasks has been provided to DPW. We anticipate this matrix being an
evergreen document of the City’s, with it being updated as new issues for DPW arise.
am
TASK 2 – REVENUE SUFFICIENCY ANALYSIS

nj
The Project Team developed a financial plan for the water and sewer system. Even though the City was
so
able to provide most of the data needed, there were certain data that could not be provided so the
Project Team developed assumptions as necessary. The financial plan indicates that with the current
ck
rates the City will just achieve necessary coverage in FY 2015 based on the assumptions made by the
Project Team. In future years, the City’s revenues will not be sufficient to meet the necessary debt
Ja
service coverage. The chart on the following page shows the revenue sufficiency challenge faced by the
water and sewer system.
The Project Team provided an overview of the financial planning model via a webinar for City staff. The
financial planning model, which is a deliverable of the task, can be used to evaluate different capital
plans, rate structures, levels of operating expenses, etc. Technical Memorandum 2, which is part of this
report, summarizes the financial results and the associated assumptions.
There is significant uncertainly regarding the level of metered consumption and magnitude of capital
and operating costs going forward for the water and sewer system. The City needs to refine the
assumptions used in the financial planning model as better information becomes available. Given the
size of the revenue shortfall in the out years, the City will need to restructure and may need to raise its
water and sewer rates, so the City should begin focusing on rate planning and structuring.
Draft Final ES-1

Exhibit 1: Expenses versus Current Revenues
om
.c
ya
la
ba
TASK 3 – BILLING SYSTEM DATA REVIEW

am
The Project Team completed an initial billing system data review. This effort commenced in late
nj
October with data acquisition and was carried out through a series of independent large-scale billing
data reviews, interviews with Information Services and Customer Service staff, and on-site account-level
so
data forensics. We first compiled historical data on consumption, billings, collections, and
customers/accounts to reconcile the billing database logs with those numbers being reported to and
ck
used by the City. Then we turned our attention to the active metering, billing, and collections data
Ja
stored in the legacy billing database.
The Project Team identified one major systemic issue, the misapplication of the minimum sewer charge,
and several inconsistencies between billing policy and data within the billing system. These included a
variety of accounts registering consumption but not being billed, exempted accounts, and vacant
accounts registering consumption. In an effort to review the advance metering infrastructure (AMI)
meter installation results, the Project Team also identified new AMI meters that were stuck at or near
zero, or were reading above a reasonable threshold. There were also large and frequent adjustments to
accounts, as well as accounts with very old accounts receivable that raised potential concerns. For each
of the ten categories of perceived inconsistencies or potential concerns, the Project Team developed a
sample of accounts for detailed review within the database’s production environment.
Draft Final ES-2

Project Team members participated in three on-site sessions to conduct detailed account-level review
and to gain a more complete understanding of existing processes. In addition to verification of potential
policy misapplication, the Project Team found many data quality concerns to be related to non-standard
practices and procedures around account adjustments, estimated reads, account exemptions, and
enforcement standards.
Overall, the goal of this task was to identify the current revenue problems based on the symptoms
found through the data analysis and detailed on-site analysis. Based on our efforts to date, we have
grouped the issues identified within three classifications:
1. Unbilled Revenues – many accounts continue to receive services and are not timely or
routinely billed, primarily due to untimely meter shut-offs and significant water theft issue.
2. Ghost Revenues – many accounts fail threshold exception edits primarily due to meter reading
errors resulting in large and frequent adjustments.
om
3. Uncollected Revenues – many accounts are delinquent and have significant accounts
receivable balances. However, the city does not have an active/proactive collections process.
.c
Going forward, the City will need assistance in remedying the problems identified. The Project Team
ya
recommends implementing a series of improved standard operating procedures and oversight
procedures for common business practices such as calculating and applying adjustments, logging and
la
verifying work orders, and collections best practices. Additionally, the Project Team recommends that
the conversion to the new CC&B billing system include measures to verify that potential process-driven
ba
failures identified within the existing system are not propagated into the new system. Testing for
conversion has begun and processes for flagging or remedying these exceptions should be included in
am
this exercise.
nj
TASK 4 –SIEMENS CONTRACT REVIEW

so
The Project Team completed an in depth review of the City’s contract with Siemens and prepared a
ck
technical memorandum (TM4) that summarizes the key terms of the contract and provides details
regarding the “guaranteed savings” that Siemens maintains will be realized by the City as a result of
Ja
work that Siemens performed under the contract. As pointed out in TM4, the majority of savings are
stipulated, which means they are assumed to occur. Since these savings are assumed to occur, the
contract does not identify a protocol for testing to determine if the savings are realized. The table and
chart below show the level of stipulated savings in Annual Period 2. There is a unique amount of
Guaranteed Savings for each annual period. We chose to show Annual Period 2 because it represents
the first year of full savings. As can be seen, the stipulated savings are 65% of the total savings. The
technical memorandum also touches on the actions that must be taken by the City in order to maximize
the benefits offered by the Siemens contract.
Draft Final ES-3

Exhibit 2: Savings from Annual Period 2
om
.c
ya
la
Once Siemens completes the projects it is performing under the contract, the City will have a state of
the art water metering system and a new customer service/billing system (Oracle CC&B) with additional
ba
capabilities. These two systems are the key drivers of the increased revenue and savings that Siemens
has promised to the City. However, as discussed in TM4, the City must first ensure that the meters and
am
billing system are properly installed and configured. Second, the City must also implement changes in
its metering and billing processes.
nj
so
To ensure the City gets the greatest value from the contract, the City needs to enhance its oversight of
the Siemens contract. In addition, the City should quantify any unexpected costs or lost revenue
ck
associated with the Siemens’ team performance under the contract.

Ja
TASK 5 – INFRASTRUCTURE MASTER PLAN DEVELOPMENT
The primary objective under this task was the development of the year 1 capital improvement plan (CIP)
for projects that may be funded using receipts from the Infrastructure Tax. The CIP must be approved
by the Commission before any of the funds from the Infrastructure Tax are utilized on projects. The
Project Team reviewed existing master plans for each of the utilities/assets (water, sewer, drainage,
streets, and bridges). Unfortunately, the City does not have recent CIPs for each utility/asset group. The
Project Team worked with City staff to prepare a draft CIP and revise the draft after comments provided
by City staff. Originally, the City was going to take the lead on this analysis, but due to limited staff
resources and time constraints, the Project Team led the effort to prepare the CIP that was presented to
the Commission in January.
Draft Final ES-4

Another component of the task was to understand the policies and procedures associated with the
revenue from the Infrastructure Tax. The Project Team met with the City’s Finance Department staff
and discussed the Infrastructure Tax. As part of this discussion, we contacted the State to gain a better
understanding of the components of the Infrastructure Tax.
TASK 6 – ADDITIONAL EFFORTS
Following the initiation of the project, the City requested the Project Team to undertake three additional
tasks.
 Moody’s Rating Evaluation Support – Moody’s, who has rated the City’s debt in the past,
requested an opportunity to update its ratings on the outstanding debt. As the process
unfolded, the City asked the Project Team to help as possible because the City does not
currently have a Financial Advisor. This assistance involved compiling and reviewing data
om
requested by Moody’s, participating in internal calls and calls with Moody’s to discuss the
situation in Jackson and the data provided, and reviewing the ratings report and helping to craft
.c
a press release.
 Capital Funding Options Analysis – The City has significant capital needs. In order to maximize
ya
the capital projects undertaken, the City needs to understand various capital funding options.
la
The Project Team provided a comprehensive overview of potential options available along with
some commentary as to the relevance to the City.
ba
 Green Infrastructure Challenge Assistance – The Infrastructure Tax is going to provide revenue
for needed drainage projects. The City is interested in a mixture of green and gray solutions to
am
overcome drainage challenges. In an effort to leverage resources and better understand the
potential capability of green infrastructure applications to solve drainage issues, the City would
nj
like to undertake a Green Infrastructure Challenge. This task involved attending meetings
related to the Challenge and producing materials, such as flyers, to publicize the Challenge.
so
ck
Ja
Draft Final ES-5

Technical Memorandum #1
Strategic Planning Issues
The Strategic Planning task was the initial task of the Project Team’s engagement to provide financial
and management consulting to the City of Jackson (City). As part of this task, members of the Project
Team participated in multiple strategy meetings and discussions with key City staff, which are
summarized in this technical memorandum. During the initial meetings, the City staff and Project Team
members brainstormed about the various challenges facing the City. Primarily, the Project Team
listened as City staff enumerated various issues. The results of these brainstorming sessions were
captured and the Project Team aggregated issues as appropriate. Ultimately, we identified 37 issues.
These issues were categorized into the following five groups.
1. Financial Management
2. Human Resources
om
3. Infrastructure Investment/CIP
4. Regulatory/Compliance
.c
5. Service
ya
After categorizing the issues, the Service category had the most issues (14), followed closely by Financial
Management (13).
la
ba
The next step was to prioritize the issues. In particular, issues were grouped as needing focus 1) within
the next six months, 2) in six to twelve months, or 3) outside of a year. The initial prioritization was
am
based on the potential impact of the issue. There were 21 issues that were prioritized in the “within the
next six months” grouping. Some issues received multiple prioritizations. The reason for multiple
nj
prioritizations for certain issues was the RFC Project Team believed the City needs to take some steps
associated with the issue immediately and some steps at a later date. A good example are two of the
so
Service issues related to Communications. Public outreach, for example, needs to occur to some extent
in the near term and the City should develop a long-term approach to public outreach regarding Public
ck
Works issues. As the City’s situation evolves, it is likely that the prioritization may need to be modified.
Ja
The Project Team also identified additional information that should be included in the matrix over time.
The City should determine the fiscal impact of addressing the issue, both the direct cost of addressing it
and the potential impact from it being addressed. For example, the fiscal impact of providing refunds to
customers outside of the City includes the cost of calculating and dispensing the refunds and the total
amount of the refunds. The City should also identify the person on City staff that is responsible for
overseeing the issue (“City Staff Lead”). Being the lead for the issue does not mean that person has to
do all the work resolving the issue, but that the person provides the oversight for the resolution of the
issue and has authority consistent with the responsibility. Thirdly, the City should identify the key
stakeholders for each issue.
The matrix is an Excel file that was originally developed in November and has been updated from time
to time. The worksheets from the current matrix are attached to this technical memorandum. The
Draft Final TM1-1

Project Team recommends that the Excel file remain a living document for the Public Works
Department. Undoubtedly, additional issues will arise and should be added to the matrix.
om
.c
ya
la
ba
am
nj
so
ck
Ja
Draft Final TM1-2

0-6 6 -12 > 12 Key Stakeholders Assigned City
Category Sub-Category Issue Description months months months Fiscal Impact Status Involved Staff lead Notes
Next permit renewal is in 2017. Addressing

this might hinge on determining in-house
resource that will coordinate. In addition,
Cost of effort (staff and possibly this can be addressed by Program
Regulatory/ Address storm water permit consultant time) and potentially Managers funded through the
Compliance Compliance requirements X higher expenses DPW staff, MDEQ Infrastructure Master Plan.
Addressing this might hinge on determining

in-house resource that will coordinate. In
Cost of effort (staff and consultant addition, this can be addressed by Program
Regulatory/ Address consent decree time) and potentially higher capital DPW staff, MDEQ, Managers funded through the
Compliance Compliance requirements X costs Program Manager Infrastructure Master Plan.
om
Identify universe of contracts that should
be assessed. Need to set a limit as to how
large a contract needs to be to focus on it.
.c
All large contracts (such as Siemens, United
Need to ensure contract compliance Cost of effort (staff and consultant Water etc.) should be addressed short
ya
Financial Contract and vendor invoice accuracy and time), but potentially offset by term. Other major contracts can be
Management Management appropriateness X X X reduced contracting costs addressed mid to long term.
Infrastructure Sales Tax Analysis - Included in current scope of work with RFC
Financial Financial policy, accounting and administration RFC team and DPW team - see executed scope for details of
la
Management Operations is weak or lacking X Ongoing staff what will be addressed
ba
Billing Issues - need to determine Cost of effort (staff and consultant Included in current scope of work with RFC
Financial Financial cause of significant under collection time), but potentially offset by RFC team and DPW team - see executed scope for details of
Management Operations relative to estimates X higher revenues Ongoing staff what will be addressed
am
The optimization of revenue collections is
Need to identify alternative revenue part of the objective of the data review that
sources to ensure no obvious Cost of effort (staff and consultant is ongoing. Identifying and evaluating
Financial Financial revenue gaps and to optimize time), but potentially offset by Partially RFC team and DPW alternative revenue sources should be a
Management Operations revenue collections X
nj X higher revenues ongoing staff focus later once things settle down a bit.
Ensure compliance with bond Included in current scope of work with RFC
so
Financial Financial requirements, specifically revenue RFC team and DPW team - see executed scope for details of
Management Operations sufficiency X Ongoing staff what will be addressed
ck
City needs to determine refunds to those

Cost of effort (staff and consultant customers charged at a higher rate than
Financial Financial Refunds to customers outside time) and cost of refunds; amount allowed by MS PSC; also need to determine
Ja
Management Operations Jackson X of refunds to be determined Amendment 2 process for "paying" refunds.
Consider conducting a water loss audit to

verify and measure magnitude of water loss
in the short-term. The audit provides the
benchmark and then allows the City to
determine progress towards improvement.
Activities that will result in reductions of
Cost of effort (staff and consultant water loss typically hinge on
time) and cost of implementing billing/metering (passive losses), upgrades
measures to reduce water loss; to linear assets or response time to water
Financial Financial Significant Water losses - approx. may result in lower operating main breaks. These strategies are typically
Management Operations 40% X X expenses in the future implemented in long term window.
Draft Final TM1-A1

Grants/City Contributions/Indirect
Cost Allocations - Need to The current flow of money between the
understand all non arms length City, DPW and the water/sewer enterprise
transactions with the city and other funds needs to be identified; an
Financial Financial city agencies to ensure no appropriate process needs to be developed
Management Operations revenue/expense gaps X and documented.
Cost of effort (staff and consultant Evaluate current debt service structure and
Financial Financial time), but potentially offset by FA, City staff, RFC discuss challenges/opportunities for future
Management Operations Optimize debt structure X reduced debt service team debt structuring
Since labor costs are the biggest
Need better handle on payroll costs - Cost of effort (staff and consultant component of operating expenses, DPW
Financial Operating OT and benefits; headcount; pay time), but potentially offset by needs to get a better handle on its payroll
om
Management Efficiency rates, etc. X reduced operating expenses costs.
Review current accounting practice
.c
that comingles operating and capital
costs. Implement change where
ya
Financial appropriate that ensures separation Cost of effort (staff and consultant The City needs to establish guidelines for
Management Policies of capital and operating costs. X time) capital versus operating costs
Inadequate financial and budget The City needs to identify and document
la
Financial policies - Need to document/update key financial policies to guide future
Management Policies where appropriate X budgeting and ratemaking.
ba
Identify and implement efforts to Cost of effort (staff time and Happy workers are more productive so
Human Resources Morale enhance employee morale X X X materials) DPW Staff DPW needs to focus on employee morale
am
Need to address key position vacancies
within the dept. There is a short-term need
Assess organizational structure to Cost of effort (staff and possibly to fill some key vacancies, but a broader
Operating ensure optimized organizational consultant time), but potentially reorganization should be undertaken once
Human Resources Efficiency outcomes X
nj X lower expenses things have settled a bit.
so
DPW should have a recruiting plan. Once
good employees are found, it is much less
Recruitment and Need to be able to recruit and retain Cost of effort (staff time and expensive to keep them than to find new
ck
Human Resources Retention high-quality staff X materials) DPW Staff ones, so retention efforts are critical.
Need to identify the level of staffing With out adequate staffing levels and
Ja
necessary to meet the required capable employees filling those roles, the
service levels and to define the City will not be able to maintain
Human Resources Staffing qualifications for this staff X DPW Staff improvements.
Cost of effort (staff time and

materials), but hopefully lower In order to have and retain high quality
Need to develop a comprehensive operating expenses with a more staff, DPW will need to identify and offer
Human Resources Training training program for DPW staff X efficient workforce DPW Staff training opportunities
It may be possible for technological
enhancements to ultimately reduce
Cost of effort (staff and possibly operating costs. Potential upgrades should
Infrastructure Operating Need to ensure technology is consultant time), but potentially be identified and evaluated in the mid to
Investment/CIP Efficiency optimally leveraged X lower expenses long term.
Draft Final TM1-A2

Included in current scope of work with RFC

Need to finalize Infrastructure Master team - see executed scope for details of
Plan, including: Service areas; Needs what will be addressed. This task is
and Costs; time frame (phase one is essential for overall financial picture and
only 1st year plan ; prioritization need to be primary priority. Phase 1 will
Infrastructure (critical, availability of funding; RFC team and DPW address immediate needs, including
Investment/CIP Capital Program mayors demo projects etc.) X Ongoing staff creating funding for PMs.
Need to finalize Infrastructure Master

Plan, including: Service areas; Needs
and Costs; time frame ( 2 - 5 year Would be addressed by Program Managers,
plan in phase two); prioritization Cost of effort (staff and consultant once brought on board. Key component of
om
Infrastructure (critical, availability of funding; time); CIP funding approaches will the year 1 Infrastructure Master Plan is
Investment/CIP Capital Program mayors demo projects etc.) X determine ultimate cost Amendment 2 getting a PM on board.
.c
Challenge is this needs to happen ASAP, but
it is not included in the RFC team scope.
ya
The long range financial plan needs to be
developed before fully implementing as we
need to know what to communicate.
However, in the short term, some general
la
Public Outreach - need to develop Cost of effort (staff and consultant good practice communication tasks could
program to effectively communicate time); level of cost is a function of be undertaken to start to build goodwill
ba
Service Communication rates and impacts X X X level of outreach with customers and stakeholders.
Goal should be to identify who the
am
stakeholders are and a plan for engaging
each one. It may make sense to engage
them together or we may want to engage
Multiple stakeholders appear to want some separately. In the short term, we
to wrestle control from DPW - need need to identify the stakeholders and their
to identify, understand and respond
nj Cost of effort (staff and consultant issues. Mid term and long term activities
timely and effectively to various time); level of cost is a function of can then be developed to address them, if
so
Service Communication Stakeholder Interests X X X level of stakeholder interaction they are not already on this list.
Identify universe of contracts that should

ck
be assessed. Need to set a limit as to how

large a contract needs to be to focus on it.
All large contracts (such as Siemens, United
Ja
Cost of effort (staff and consultant Water etc.) should be addressed short
Contract Need to assess and ensure contract time), but potentially offset by term. Other major contracts can be
Service Management performance X X X reduced operating expenses addressed mid to long term.
Cost of effort (staff and consultant Included in current scope of work with RFC
Contract Siemens - WW facilities; contracted time), but potentially offset by RFC team and DPW team - see executed scope for details of
Service Operations operations - ensure oversight X reduced operating expenses Ongoing staff what will be addressed
Initial thinking was the City needs to best
manage contract until it ends in about a
year, at which time the City would put
contract out to bid. However, it appears
Cost of effort (staff and consultant that new and pressing issues related to
Contract United Water contract (contracted time), but potentially offset by contract may require more immediate
Service Operations WWT operator) - ensure oversight X reduced operating expenses action.
Draft Final TM1-A3

Included in current scope of work with RFC
Contract Siemens - Operating plan post RFC team and DPW team - see executed scope for details of
Service Operations Siemens X Ongoing staff what will be addressed
Chemicals and electricity are typically the

Need to ensure cost of Production is largest costs behind labor, so these costs
in line with industry- treatment, Cost of effort (staff and consultant need to be evaluated. One option is to
Operating collection, distribution (chemicals, time), but potentially offset by benchmark production costs with those of
Service Efficiency electricity, people) X reduced operating expenses peer utilities.
Cost of effort (staff and consultant

Ensure key operating policies and time), but potentially offset by DPW should develop operating policies to
Service Policies procedures are in place X reduced operating expenses help minimize operating expenses.
om
Cost of effort (staff and possibly DPW should move towards budgeting
consultant time) and potentially based on level of service. In order to do
higher expenses to meet DPW Staff and City this, it is necessary to establish the level of
.c
Service Service Levels Define appropriate level of service X appropriate service levels Council service the community desires.
Review audit performed by WRUA and
ya
Issues with agreement - West Rankin Cost of discussions (staff and determine what needs to be done to
Wholesale Utility Authority (WRUA); need to consultant time) and potentially RFC Team, DPW address the issues and the potential impact
Service Agreements address United Water audit issue X refunds and less future revenue Ongoing Staff, and WRUA on the City
There appear to be a number of issues
la
Cost of discussions (staff and DPW staff and between the City and its wholesale
Financial Wholesale Potential issue with rates currently possibly consultant time) and Wholesale customers. The first issue that should be
ba
Management Agreements assessed to utilities outside city limit X potentially less revenue Customers evaluated is the rate issue.
am
Cost of discussions (staff and
possibly consultant time) and Identifying the issues may be short to mid
Wholesale Issues with agreement - City of potentially less revenue or higher DPW staff and City term, but typical strategies to address
Service Agreements Ridgeland X nj expenses of Ridgeland would probably be long term

so
Wholesale Issues with agreement - Madison potentially less revenue or higher DPW staff and term, but typical strategies to address
Service Agreements County X expenses Madison County would probably be long term
ck

Ja
Wholesale potentially less revenue or higher DPW staff and term, but typical strategies to address
Service Agreements Issues with agreement - Hinds County X expenses Hinds County would probably be long term
Wholesale potentially less revenue or higher DPW staff and City term, but typical strategies to address
Service Agreements Issues with agreement - City of Byram X expenses of Byram would probably be long term
Draft Final TM1-A4

Revenue Sufficiency Analysis
The purpose of the Revenue Sufficiency Analysis was for the Project Team to develop a financial
planning model for the City of Jackson’s (City) water and sewer system (System) which could be used 1)
to evaluate the System’s ability to meet debt service coverage under current assumptions and 2) to
address the System’s potential inability to meet debt service coverage in the future. The analysis was
required because the City did not meet its fiscal year (FY) 2013 rate covenant obligations associated with
the issuance of revenue bonds. The revenue sufficiency analysis also addressed the deficiencies noted
by Moody’s in their recent (November 2014) downgrade of the City’s bond rating, which were as
follows:
 Inability to meet debt service coverage requirements in FY 2013;
 Undeveloped financial plan;
om
 Declining unrestricted cash;
 Limited rate raising history; and
.c
 Sizable consent decree which requires additional debt leveraging.
BACKGROUND
ya
la
In 2013, the City issued revenue bonds (FY 2013 Bonds) for approximately $91 million to cover projects
ba
to be performed by Siemens Industry under a performance contract, specifically for the implementation
of advanced meter infrastructure (AMI), improvements to the O.B. Curtis and J.H. Fewell water
am
treatment plants, and replacement of several major sewer collection lines. At the time the FY 2013
bonds were issued, it was projected that the City would be able to meet the obligations of the FY 2013
nj
revenues bonds through operational savings and revenue enhancements stipulated by the performance
contract (totaling approximately $7.8 million per year once all the projects were finished and
so
operational) combined with moderate water and sewer rate increases. However, due to various
ck
circumstances, as explained later in this memorandum, the City was unable to meet the obligations of
the FY 2013 bonds.
Ja
STUDY APPROACH
In order to provide a current financial picture of the Water and Sewer System, the City engaged the
Project Team in the fall of 2014 to develop a comprehensive revenue sufficiency study. To begin the
study, the Project Team requested and the City provided the following information:
 FY 2015 detailed operating budgets for both the water and sewer utilities;
 Historic data on operating expenses and revenues;
 Detailed amortization schedules for all outstanding debt including, revenue bonds, general
obligation bonds (GO), state revolving fund (SRF) loans, Mississippi Development Authority
(MDA) loans, and capital leases;
 Comprehensive annual financial reports;
Draft Final TM2-1

 Official statements regarding bond issues;
 Historic customer information including number of customers by location and billable water and
sewer flow;
 Historic data on adjustments to billable water to determine collection rates;
 Overall consent decree project costs;
 Capital improvement plan for FY 2015; and
 Infrastructure sales tax revenue available to be applied towards the System’s anticipated capital
improvement plan.
The Project Team reviewed this data in detail and developed a financial planning model using the
information obtained above. The Project Team has solicited additional information and clarification of
information from City staff several times throughout the development of the financial planning model.
RFC also made several assumptions, as explained later in this memorandum that impact the Revenue
Sufficiency Analysis.
om
Study Assumptions and Results
.c
ya
The Project Team projected the annual cash needs (revenue requirements) of the system for FY 2015
through FY 2020 and compared these costs to the projected revenues under existing rates. RFC also
la
calculated the debt service coverage ratios in each year to determine if the existing rates are sufficient
to meet the debt service coverage requirements. The assumptions used in projecting the annual
ba
revenue requirements, revenues, and debt service coverage are explained below.
am
Revenue Requirements Assumptions

The Project Team identified the System’s annual cash needs, or revenue requirements, which include
nj
operations and maintenance (O&M) expenses, existing and proposed debt service, and cash-funded
so
capital projects, which are explained below:

ck
 The detailed FY 2015 O&M budget was used to project future O&M expenses as follows:
 Most line items were escalated by 3% based on input from City staff regarding
Ja
anticipated future costs.

 The Department of Public Works is conducting a salary study to determine the impact of
minimum wage rates. Because this study was not finalized as of the date of this
memorandum, the Project Team has assumed personnel costs will increase by 10% in FY
2016.
 In order for the City to provide the desired level of water and sewer service to its
customers, the City will incur additional O&M costs to eliminate water loss, address
reliability issues, and reduce the volume of leaks and breaks. Therefore, the Project
Team assumed O&M costs would increase by $250,000 in FY 2016 for water and for
sewer and then an additional $250,000 in FY 2017 and beyond (totaling $500,000 each
for water and sewer by FY 2017) which represent increased operating and maintenance
costs.
Draft Final TM2-2

 It should be noted that RFC has not assumed any operational savings from the Siemens
performance contract in its forecast. (See explanation provided later in this
memorandum).
 Debt service schedules were used to identify the total annual debt service to be paid in FY 2015
– FY 2020.
 Because the Public Works Department is currently developing a five-year capital program, the
Project Team had to estimate the capital improvement plan (CIP) and the funding sources as
follows:
 The Project Team used the overall consent decree ($400 million over 17 years) to
estimate a level of sewer projects ($23.5 million per year).
 The Project Team used the recently developed CIP for FY 2015 in which it identified
approximately $8 million in water projects. This amount was assumed to continue into
the future but will more than likely increase as the City addresses water loss issues,
reliability issues, and leaks and breaks.
om
 RFC determined funding sources, which are estimates until the City engages a Financial
Advisor. The funding sources include a portion of the Infrastructure Sales Tax, GO
.c
Bonds, Mississippi Development Loans (MDA), and State Revolving Fund Loans (SRF). It
should be noted the City does not believe that SRF Funds will be available until future
years (FY 2019 – 2020).
ya
la
 The City currently does not have an internal target regarding the level of the CIP that
should be cash funded versus debt funded. Cash funding a portion of the CIP allows for
ba
stronger debt service coverage. RFC is recommending an internal target of cash

financing 15% of the CIP by FY 2020, which is incorporated into the study results.
am
Revenue Assumptions
nj
The Project Team estimated annual revenues including revenues from retail customers, wholesale
customers, and miscellaneous services, as follows:
so
ck
 To estimate revenues from retail customers, RFC obtained detailed customer billing
information for the past two years on each customer. However, due to the large discrepancy
Ja
between the City’s billable data and the amount of money that is actually received, RFC had to
use data on actual revenues to back into billable flow estimates. The City also provided RFC
with information on the total dollar amount of adjustments given in each year. RFC used this
information to estimate the City’s amount of billable flow and then the percent of uncollected
revenues (approximately 12%). RFC has assumed that the amount of uncollected revenues will
improve over the next few years as adjustments resulting from the meter change-out and
other reasons decrease, as the City begins to enforce cutting-off customers for non-payment,
etc. It is assumed that the City will achieve an uncollectable rate of 5% by FY 2020. (The
industry average is 1%). For now, RFC has assumed that there will not be an increase in billable
flow resulting from the Siemens performance contract (improved meter accuracy), nor has RFC
assumed any increase in the number of water/sewer customers or growth in water/sewer
flow. The industry trend is declining per capita consumption, which has been assumed to
offset any increase from enhanced meter accuracy. Even though we believe this is a
Draft Final TM2-3

conservative assumption, given the financial situation of the System, RFC believes it is prudent
to make conservative assumptions at this time.
 RFC estimated revenues from sewer wholesale customers by using the existing methodology of
each customer paying a portion of O&M, debt service, and capital costs based on their
proportion of sewer flow.
 RFC incorporated miscellaneous revenues from such items as cut-off fees, returned check fees,
service connections, interest earnings, etc. and is assuming these revenues will remain
constant into the future.
Debt Service Coverage Calculation Assumptions

The Project Team calculated the debt service coverage for each year in the forecast. The debt service
coverage calculation is a two-pronged test which is as follows:
om
The greater of:
1) 120% (1.20) of annual debt service on Revenue Bonds +/- rate stabilization, OR
2) 100% (1.00) of sum of:
.c
a) Annual debt service on Revenue bonds and subordinate debt (GO, SRF, MDA,
ya
capital leases)
b) amounts to paid during year for Debt Service Reserve System and the
la
Contingent System
c) any other charges or liens payable out of Revenues during the fiscal year not
ba
otherwise provided in this subsection

am
Because of the two-pronged test, “the greater of” results in the debt service coverage requirement
being 100% of total debt (more than just revenue bonds) plus amounts required to adequately fund the
nj
Debt Service Reserve System and the Contingent System. The City currently does not have an internal
target for its debt service coverage ratio. RFC is recommending an internal target of debt service
so
coverage of 1.10 of total debt (revenue bonds, GO bonds, SRF loans, MDA loans, and capital leases),
ck
which is incorporated into the financial planning model.

Ja
Financial Plan Results

The City will continue to be in a detrimental financial position if current water and sewer rates remain
unchanged. As shown in Exhibit 1, the current revenues are barely able to meet the revenue and debt
service coverage requirements of the System in FY 2015 and debt service coverage is compromised in
future years.
Draft Final TM2-4

Exhibit 1: Expenses versus Current Revenues
om
.c
ya
la
ba
In order to meet the revenue requirements of the water and sewer utility, the City will need to
implement rate adjustments. However, the level of the rate adjustments will be impacted by the
am
following:
 Level of operational savings identified in the Siemens performance contract that is actually
nj
achieved.
 Ability of the City to achieve the revenue enhancements identified in the Siemens performance
so
contract resulting from increased meter accuracy.

 Ability of the City to increase its collection efforts by addressing the number of adjustments
ck
given to customers, enforcing the cut-off policy, etc.

Ja
 Modifications to the rate structure based on cost of service principles to enhance equity and
address affordability concerns.
While the level of rate adjustments will vary based on the factors listed above, the Project Team
recommends that the City set rates such that the City can achieve sufficient debt service coverage.
Exhibit 2 shows an example of proposed debt service coverage targets in each fiscal year that would
allow the City to meet the cash needs of the utility, which include O&M costs, capital costs, and debt
service targets.
Draft Final TM2-5

Exhibit 2: Debt Service Coverage under Existing Rates and Proposed Target
om
.c
ya
la
ba
COMPARISON OF STUDY RESULTS TO ENGINEER’S PROJECTIONS

am
As mentioned previously, at the time the FY 2013 bonds were issued, it was projected that the City
would be able to meet the obligations of the FY 2013 revenues bonds through operational savings and
nj
revenue enhancements stipulated by the performance contract and moderate rate increases. Appendix
so
J of the City’s official statement (OS) for the FY 2013 Bonds included a report titled Independent
Consulting Engineer’s Report (Engineer’s Report) which summarized the assumptions of the operational
ck
savings and revenue enhancements, and provided a projection of cash flow for FY 2013 through FY
2017. The following table is a replication of the table provided in the Engineer’s Report. It should be
Ja
noted these represent 87% of the savings/enhancements to allow for unforeseen events (or 13%
variability), as explained in the Engineer’s Report. As shown in Exhibit 3, once the projects were fully
completed and operational, it was estimated that the City would achieve operational savings and
revenue enhancements of approximately $7.8 million per year.
Draft Final TM2-6

Exhibit 3: Replication from Engineer’s Report
Timing Event Impact

End of FY 2013 Most large meters installed (all meters Large meter and deferred
to be installed by the end of December maintenance savings of $1.5 million
2013) by the end of FY 2013
End of FY 2014 All small meters installed (1/2 years Large/small meter and deferred
reduction in deferred maintenance maintenance savings of $3.4 million
expense assumed) in FY 2014
End of FY 2015 All meters installed (full year of Large/small meter and deferred
deferred maintenance expenses maintenance savings of $5.6 million
assumed) in FY 2015
End of FY 2016 System upgrades complete (full year of Large/small meter and deferred
deferred maintenance and operations maintenance and operational savings
expense assumed) of $7.7 million in FY 2016
om
End of FY 2017 Fully operational system Large/small meter and deferred
maintenance and operational savings
of $7.8 million in FY 2017
.c
ya
Exhibit 3 shows the anticipated timing and the magnitude of the operational savings and revenue
enhancements. In addition to these assumptions, the cash flow analysis in the Engineer’s Report also
la
assumed the following:
ba
 Debt service coverage of 1.20 in each fiscal year, where debt service coverage is defined as 1.20
of annual revenue bond debt
am
 Increases in volumetric water rates of approximately 4% in FY 2015, 4.6% in FY 2016, 9.3% in FY

2017
 Increases in volumetric sewer rates of approximately 2.8% in FY 2015, 6.3% in FY 2016, 2.1% in
nj
FY 2017
so
While the Engineer’s Report demonstrated the City’s ability to meet the obligations of the FY 2013
ck
bonds, in actuality the City failed to meet its debt service coverage in FY 2013. Recognizing its inability
to meet coverage, the City implemented a rate increase (that went into effect on November 8, 2013) of
Ja
29% for the volumetric component of the water rate structure and a rate increase of 108% for both the
fixed and volumetric components of the sewer rate structure so that the City could meet its debt service
obligations in FY 2014. As of the date of this technical memorandum, the audit had not been completed
to determine if these rate increases were sufficient for the City to meet coverage for FY 2014. However,
based on RFC’s preliminary analysis, it appears the City may meet its coverage requirement for FY 2014
and FY 2015 but not in future years. Nonetheless, Moody’s downgraded the City’s bond rating from A1
to A2 in November 2014 after reviewing the City’s financial information.
The disparity in the cash flow analysis in the Engineer’s Report versus the updated projections
developed by RFC is explained below and is also shown in the chart that follows the explanation.
 The timing of the project has varied significantly from the project schedule. For example, as of
October 2014, only 30% of the meters had been replaced which is in contrast to the initial
Draft Final TM2-7

schedule of having all meters replaced by the end of FY 2014. As a result, this has limited the
City’s ability to realize operational savings and revenue enhancements in FY 2013, FY 2014, and
will continue into the future until the project is completed.
 The debt coverage calculation in the original projections exactly met the debt coverage
requirement of 1.20 (of the annual debt service on revenue bonds only). This means that there
was no room for any variation in any of the assumptions (excluding the 13% allowed variability
in the revenue enhancements or operating savings that was already incorporated in the
calculation). As explained previously in this technical memorandum, the debt service calculation
is a two-pronged test. RFC’s interpretation of the rate covenant is a more stringent test than
1.20 on senior debt service.
 The original projection escalated operating costs for water at 1% and 3% for sewer. In actuality,
the operating costs have and will likely continue to increase at much higher levels due to:
 Higher chemical costs for water resulting from the well system shutdown and reliance
on the surface water treatment plant which requires more chemicals than the well
om
system.
 Higher chemical costs for sewer due to having maximized current treatment of plant
.c
discharge waters. In addition, the Siemens’ contract will repair Centrifuge #1 which
will put the discharge plant at full capacity once again. This increase in dried sludge
ya
output will cause an increase in the amount of chemicals being used.
la
 Higher electricity costs due to the surface water plant having to pump water further
since the well system is shut down.
ba
 Higher cost of operation at the new Presidential Hills facility.

 Additional staff needed to provide necessary level of water and sewer service.
am
 The original projections did not incorporate any additional debt resulting from the City’s consent
decree or capital improvement projects related to water system replacement projects.
nj
 The rate increases that were implemented (even at the much higher levels than under the
original projections) have not resulted in revenues increasing in proportion to the rate increases
so
due to a high level of adjustments to water use (some of which is attributed to the meters being
changed out and not registering correctly) and uncollectible accounts.
ck
Ja
Exhibit 4: Comparison of Engineer’s Projections to RFC’s Projections
RFC's Revenue Sufficiency Analysis Engineer's Report - Projected Cash Flow

CAFR (1) Unaudited (2) Budget (3) Budget (3) Budget (3) Projected
FY 2013 FY 2014 FY 2015 FY 2016 FY 2017 FY 2013 FY 2014 FY 2015 FY 2016 FY 2017
Total Revenues $ 44,186,000 $ 68,431,191 $ 68,431,191 $ 68,431,191 $ 68,431,191 (1) $ 51,492,722 $ 53,008,927 $ 55,735,686 $ 57,969,391 $ 60,664,138
Less: Total O&M Expenses $ (33,187,000) $ (39,203,069) $ (45,229,699) $ (47,582,084) $ (49,788,878) (2) $ (37,827,477) $ (38,200,280) $ (38,067,595) $ (36,823,253) $ (37,593,403)
Revenues Available for Debt Service $ 10,999,000 $ 29,228,122 $ 23,201,492 $ 20,849,107 $ 18,642,313 $ 13,665,245 $ 14,808,647 $ 17,668,091 $ 21,146,138 $ 23,070,735
Revenue Bond Debt Service $ 12,641,000 $ 12,641,000 $ 14,986,710 $ 17,925,216 $ 19,479,741 $ 11,424,804 $ 11,738,238 $ 14,667,492 $ 17,605,999 $ 19,160,524
Debt Service Coverage on Rev. Bonds 0.87 2.31 1.55 1.16 0.96 1.20 1.26 1.20 1.20 1.20
Must be 1.20
Total Debt Service $ 18,382,313 $ 23,467,072 $ 22,778,731 $ 26,222,279 $ 28,789,563
Debt Service Coverage on Total Debt 0.60 1.25 1.02 0.80 0.65
Must be 1.00 0.60 1.25 1.02 0.80 0.65
(1) From the FY 2013 Comprehensive Financial Annual Report (page 131), with the (1) Includes enhanced revenues from performance contract and rate increases of
exception of the revenue bond figure which is net of capitalized interest. It should be 4%, 4.6%, and 9.3% for water and rate increases of 2.8%, 6.3%, and 2.1 % for
noted that FY 2012 revenues were $48.2 million. sewer in FY 2015, FY 2016, and FY 2017, respectively.
(2) Based on unaudited information provided by City staff and includes rate (2) Includes operational savings from the performance contract.
increases implemented in November 2013 of 29% for water and 108% for sewer.
(3) Based on FY 2015 budget provided by City staff but assumes no rate adjustments.
Draft Final TM2-8

NEXT STEPS
 The Project Team recommends that City staff continue to use the revenue sufficiency model
developed by RFC as a financial planning tool and incorporate updated information into the
model as it becomes available so that debt coverage ratios can be monitored and met in future
years, and so that rate adjustments can be identified over a long-term planning period to avoid
rate shock. The financial planning model should be updated annually to reflect any operational
savings and revenue enhancements from the Siemens performance contract, other factors that
affect operational expenses such as the City’s ability to reduce water loss, etc., demand
projections that are impacted by weather, collection efforts that are impacted by adjustments
and cut-off policies, etc.
 The Project Team also recommends that the City perform a comprehensive cost of service
analysis to determine the optimal rate structure that will allow the City to balance rate
om
adjustments with affordability and equity concerns. For example, the City could have a different
rate for each customer class (residential, commercial, etc.), or tiers for residential customers
that could include a lifeline rate for those customers with low water use. The cost of service
.c
analysis could also determine the level of costs to be recovered from the fixed component of
ya
the rate structure versus the volumetric component that varies by water/sewer flow.
Depending on the results of the cost of service analysis, the rate adjustments would impact
la
customers differently. A comprehensive cost of service analysis would allow the City to balance
rate adjustments with its rate structure pricing objectives including, but not limited to,
ba
affordability and equity.

am
nj
so
ck
Ja
Draft Final TM2-9

Billing System Data Review
OBJECTIVE
In recent months, the City of Jackson (City) has recognized a lower than expected level of water and
sewer revenue in light of the recent rate increases. The collection rate fell significantly between 2012
and 2014, and current revenues are materially below expected totals. The Project Team was contracted
to perform a variety of financial support roles, one of which was to identify patterns in billing and
collections that may explain low revenue figures using active and historical billing system data. As
defined in the project scope regarding potential billing data inaccuracies, the Project Team sought to
determine:
 Who is affected?
om
 How it is occurring?
 Why it is occurring?
 How much revenue is potentially at stake?
.c
ya
The approach to this was to examine the billing data from meter readings to incoming payments to
identify any process disconnects, data quality concerns, incorrect application of rates or billing policy,
la
fraudulent activity, or other drivers behind the reduction in revenue. Armed with that data, the Project
ba
Team sought to find eligible revenue sources wherever possible by identifying accounts not being billed
when they should be or not being charged the correct amount.
am
The Project Team had the additional objectives of supporting financial planning with live billing data and
determining the extent (if at all) of incorrect billing to a group of customers outside the City.
nj
so
As the City goes through conversion to automated metering infrastructure (AMI) and a new billing
system, identification of any issues is critical to limit their perpetuation into the new environment and to
ck
resolve any outstanding revenue sources before those data are potentially lost in the transition. This
transition process has been temporarily suspended, but is expected to continue. Even with an extended
Ja
delay, improvements to data quality and data management processes can increase billing and
collections in the current environment as well.
PROCESS
Data Acquisition
This effort commenced in late October with data acquisition and interviews with Information Services,
utility management, and Customer Service staff. The Project Team obtained an extract copy, dated
October 17, 2014, of the legacy billing system to be restored in our Oracle environment. We also
obtained daily history files generated on a nightly basis as a record of all changes to the billing data from
1999 through November 7, 2014. To support review of these data, the Project Team received
documentation including lookup codes for billing system data, entity relationship diagrams, and SQL
scripts used by staff to perform regular and ad hoc queries against the database.
Draft Final TM3-1

During this initial visit, the Project Team held discussions with Information Systems and Customer
Service staff on relevant billing policy information and data structure, and learned of some known
problems with data and processes as well as some problems already being addressed. These included:
1. Hung accounts due to outstanding work orders;
2. New AMI meters that are stuck;
3. New AMI meters causing dramatic billing increases;
4. Accounts registering consumption with no present occupants; and
5. G/E code accounts that are exempt from enforcement actions.
Data Structure
The legacy billing system includes a number of tables, but only about 10 that are used to manage day-to-
day customer, consumption, and billing data. Of these, the majority of available billing information is
om
stored in the Meter Detail table, with 1,027,980 records representing the last 12 reads and bills for every
meter. In our extract, this table included read and bill information from October 2012 to October 2014
for most customers. The Accounts and Meters tables each have 85,665 records. Meters are related with
.c
accounts and customers through a series of cross-reference tables.
ya
The daily history files contain a wealth of information including logs of every bill generated, every
la
adjustment made, every payment made, and most user edits. For reference, the following table shows
ba
the number of records in each of the following categories, representing transactions from 1999 through
November 7, 2014.
am
Table 1. Amount of data contained in daily history files

nj
TABLE NUMBER OF RECORDS

so
ADJUSTMENTS 138,554
CHECK PAYMENTS 3,184,815
ck
BANK DRAFT PAYMENTS 301,415

CASHIERING PAYMENTS 1,572,836
Ja
USER EDITS 3,487,316

BILLS 5,546,163
High Level Review

To gain a high-level understanding of utility finances, the Project Team used daily history file data
including every bill generated and adjustment recorded. These data were used to drive the review as it
extended back to 1999 and gives a rich history, whereas only two years are typically available in the live
billing system data due to constraints on the system. With the daily histories of bills generated and
adjustment activities, the Project Team pieced together long-term trends on consumption, billings,
collections, customers/accounts. From this perspective, the Project Team was able to visualize the
reduced revenue, seeking an obvious driver for the problem.
Draft Final TM3-2

These data showed that the number of accounts has been decreasing slightly over time, though
consumption is holding nearly steady. Billings, before any adjustments are applied to accounts, have
been rising steadily over time and have steep increases aligned with the timing of rate increases. At the
same time, adjustments have followed a similar trend, increasing drastically even normalized for rates.
These adjustments are mostly related to water charges rather than sewer, sanitary, or meter charges.
As a result, adjusted billings have not increased as much as expected (see Figure 1).
Figure 1. Total Unadjusted Charges, Adjustments, and Adjusted Charges by Year
$120.00
$100.00
om
$80.00
.c
$60.00
ya
$40.00 la
$20.00
ba
$-
am
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Billings Adjustments Adjusted Bills

nj
so
Payments, though increasing, have not been tracking with expectations for the past few years. This is
especially evident with cashier payments. While it is understandable that the proportion of cashier
ck
payments may decrease over time as more customers use alternate payment options, the revenue from
these other options, mailed checks and bank drafts, has not made up the difference. As a result,
Ja
outstanding balances are on the rise and the collection rate is decreasing. Generally, increased
adjustment activity and decreased collections are the major drivers for the reduced revenue stream.
These figures are summarized in the table below.
Draft Final TM3-3

Table 2. Summary of High Level Data
#
% ADJUSTED % BILLED
YEAR BILLINGS ADJUSTMENTS ADJ BILLS CHECKS DRAFTS CASH RECEIPTS COLL ACCTS
1999 23,029,867.37 (2,326,493.71) 10% 20,703,373.66 14,383,678.49 $836,959.71 4,226,444.30 19,447,082.50 94% 65,695
2000 48,247,048.87 (5,445,588.17) 11% 42,801,460.70 29,805,917.52 $1,867,040.39 8,915,791.76 40,588,749.67 95% 67,608
2001 38,402,584.45 (4,497,069.42) 12% 33,905,515.03 23,591,948.94 $1,401,105.28 7,797,002.50 32,790,056.72 97% 66,952
om
2002 56,753,862.31 (5,600,952.03) 10% 51,152,910.28 36,934,634.94 $1,999,840.60 12,339,446.37 51,273,921.91 100% 67,100
2003 56,587,856.06 (8,499,363.39) 15% 48,088,492.67 33,166,028.51 $1,812,692.99 11,540,326.87 46,519,048.37 97% 66,562
2004 56,581,223.32 (7,087,090.84) 13% 49,494,132.48 32,506,980.96 $1,954,803.87 13,201,977.92 47,663,762.75 96% 66,308
.c
2005 63,710,498.81 (9,073,461.72) 14% 54,637,037.09 35,234,866.46 $2,115,425.67 12,876,349.74 50,226,641.87 92% 66,047
ya
2006 61,422,308.90 (8,797,588.37) 14% 52,624,720.53 33,866,342.29 $2,071,952.58 16,087,833.33 52,026,128.20 99% 65,704
2007 60,481,050.54 (8,259,673.39) 14% 52,221,377.15 31,964,535.59 $2,041,538.74 17,530,588.90 51,536,663.23 99% 64,718
la
2008 67,916,162.21 (11,497,746.82) 17% 56,418,415.39 32,768,787.23 $2,111,619.58 16,956,346.66 51,836,753.47 92% 64,237
2009 66,261,029.90 (13,871,447.53) 21% 52,389,582.37 29,398,409.67 $1,836,450.16 17,854,046.15 49,088,905.98 94% 64,382
ba
2010 65,101,024.67 (11,088,840.49) 17% 54,012,184.18 32,016,889.07 $1,985,247.46 17,551,258.58 51,553,395.11 95% 64,025
2011 71,323,964.23 (14,109,562.55) 20% 57,214,401.68 36,677,896.50 $2,146,939.19 16,623,617.54 55,448,453.23 97% 63,548
am
2012 73,038,101.34 (16,566,067.84) 23% 56,472,033.50 36,003,024.74 $2,026,614.66 15,291,882.30 53,321,521.70 94% 62,595
2013 74,560,102.15 (14,932,747.34) 20% 59,627,354.81
nj 39,001,206.03 $2,192,379.71 13,045,818.91 54,239,404.65 91% 61,900
THRU NOV
95,777,131.41 (26,773,241.38) 28% 69,003,890.03 45,061,498.44 $2,676,389.43 12,903,999.69 60,641,887.56 88% 62,819
7, 2014
so
2014 PROJ. 110,143,701.12 (30,789,227.59) 28% 79,354,473.53 51,820,723.21 3,077,847.84 14,839,599.64 69,738,170.69 88% 62,819
ck
Ja
Draft Final TM3-4

Bulk Billing Data Review
A bulk analysis of billing data in the live billing system was performed to supplement this finding with
specific data quality concerns that could be leading to unbilled accounts or accounts requiring
adjustments. Based on known rates, billing policies, and data structure, the Project Team used customer
and consumption data in the billing system as of October 2014 to replicate account water and sewer
charges before and after the November 2013 rate change. This exercise resulted in the identification of
categories of accounts where the customer charges were not as expected, where bills seemed to be
going out or not going out unexpectedly, and where poorly maintained customer information could
result in unenforceable circumstances. These circumstances could be symptoms of a systematic
misapplication of billing policy or simply of poor data quality and limited quality control measures. The
results of this analysis are described further below.
Field Audit
Based on the bulk billing data review, RFC developed sample data for further investigation by project
om
partners, IBS. This investigation took place on site over three separate visits, during which IBS staff has
worked with City Information Systems and Customer Service staff to research individual accounts and
.c
determine whether the exceptions noted during bulk review were justified, or whether other similarities
exist between accounts exhibiting the same unexpected circumstances.
ya
Field visits were conducted on December 9th and December 18th and 19th, 2014, and the week of January
la
5th, 2015. The findings from these visits are summarized below and a detailed report is included as
ba
Appendix A.
am
FINDINGS
nj
Billing Data Accuracies

The Project Team could replicate the vast majority (approximately 97%) of water charges to accounts
so
within the City. Many of those that were not replicable were not receiving bills, which is addressed
below. This is not to say that payments are made to all correctly charged accounts, but that the Project
ck
Team found the billing system data related to water charges to be mostly accurate.
Ja
Billing Data Inaccuracies

Data Validity
While reviewing the data in bulk, the Project Team found many instances of invalid entries throughout
most tables and fields. For example, there are accounts within the City with both inside- and outside-city
flags in different fields. Rates are determined based on a meter’s location (inside vs. >1 mile outside the
city) and having conflicting information on the same account could confuse the calculation process,
especially as it is converted to the new billing system.
Similarly, there are entries that make the billing policy difficult to accurately implement. Residential
properties with multiple units are supposed to be billed per unit. In the legacy system, there are two
fields that denote this set-up. First, the Apartment Code should be “0” for single family properties (one
unit) or “A” for multiple units. A subsequent field is supposed to contain the number of units where the
Draft Final TM3-5

Apartment Code is “A.” The Project Team found 38 other unique values other than other than “0” or “A”
included in the Apartment Code field of 173 accounts. The Project Team also found 13 instances where
more than one unit was noted with a “0” single family residential property, and 386 instances where “A”
multi-unit properties were listed with 0 or 1 as the number of units. All of these are invalid combinations
that could result in the miscalculation of per unit water and sewer charges.
Across the board, there were invalid data that hindered bill replication by the Project Team, but more
importantly caused concern about data maintenance and data quality as it relates to correct bill
calculation at the City.
Adjustments
Given that adjustments appeared to play a large role in the reduction of revenue in 2014, two groups of
accounts were identified to examine recent adjustments. All large adjustments, defined as a single credit
or debit of more than $400, and all frequent adjustments, defined as four or more adjustment instances,
om
over the past two years were pulled out and sampled for further review. As of November 2014, there
were 3,391 large adjustments totaling $(41,204,999) and 141 accounts with many adjustments, totaling
.c
$(2,562,740). The project team found no systematic evidence of accounts being adjusted by values
greater than the account balance, except in cases where payments had already been made.
ya
From the field audit, the root cause(s) of the adjustments issue appears to be mostly human errors from
la
meter readers. The readings are fixed through a re-compute system process, but before this occurs they
ba
are already registered in the system incorrectly, causing an adjustment to appear on the record. Another
root cause is meter leaks, which are addressed by a manual calculation to determine a corrected bill
am
based on past consumption. In this case, only a monetary adjustment is made, and no consumption
adjustment is done in the system.
nj
The City does have a leak adjustment policy and a standard process for meter error adjustments. All the
so
adjustments sampled were processed in accordance with policy as described and were properly
authorized. It appears that adjustments of any level are signed off by supervisors before being
ck
committed to the system.

Ja
The Project Team noticed that in many cases, several adjustments were made to an account at a single
time, seemingly as a series of attempts to get the calculation to end up at a desired result. That, paired
with reported supervisory authorization for each adjustment, suggests that more stringent supervisory
review and more carefully calculated adjustments should reduce this value. The billing system does not
have any validation rules or checks on an adjustment before it is committed. Going forward, these are
highly recommended. The City needs to instate a process or improve existing processes to address data
inaccuracies before they go into the billing system.
Many Adjustments
For the field audit, the team identified a sample of 159 accounts with more than four adjustments over
the most recent two year time period for further analysis. Of those, the team received additional work
order information for 52 (33%), representing adjustments of $1,011,722. 44 (85%), which were
Draft Final TM3-6

associated with work orders to address meter reading errors, and 8 (15%), which were associated with
work orders for meter leaks.
Large Adjustments
For the field audit, the team identified a sample of 61 accounts with more than four adjustments over
the most recent two-year time period for further analysis. Of those, the team received additional work
order information for 30 (49%), representing adjustments of $1,367,876. 16 (60%), which were
associated with work orders to address meter reading errors, and 2 (7%), which were associated with
work orders for meter leaks.
AMI Meters
City staff mentioned that the installation of new AMI meters had been causing a number of billing issues
to be reported by customers. Primarily, the new meters either were causing no consumption to be
billed, or dramatic increases to be caused in bills. These two issues were examined in the billing data and
om
subsets of accounts were identified for detailed examination, as explained below.
.c
Stuck AMI meters
The issue of stuck AMI meters has been addressed in part by the meter installer through improved
ya
quality control measures. However, more instances than were acknowledged by the installer were
found in the billing data. The Project Team identified a set of 283 AMI meters linked to active accounts
la
with a history of steady consumption that drops drastically (to or near 0) after the meter is replaced.
ba
During the field audit, the Project Team found that the standard for stuck meter resolution is 8 weeks,
am
though no policy or service level agreement was provided. Our review revealed that about 90% meet
this standard based on service order notes in the sample group. As of the date of the audit, only 2
nj
accounts (10%) on the sample list were still unresolved. While this process seems to be improving, the
service order process within the legacy system is open-ended and there are many opportunities for
so
service orders to fall through the cracks. The service order entry and follow-up process is described
more fully below.
ck
Reading above Threshold

Ja
Another claim regarding the new AMI meters is that some appear to be miscalibrated or reading in the
wrong unit (gallons instead of cubic feet). Theorizing that these problems could be teased from the data,
the Project Team developed a list of 3,371 new meters whose readings after being replaced with AMI
meters were two standard deviations above mean consumption prior to the replacement.
During the January 2015 field audit, the Project Team discovered that DPW was experiencing a high rate
threshold exceptions, averaging approximately 53% of all billed accounts needing review. The field
auditors sampled 30 accounts. Twenty-one (70%) of the exceptions appear to be triggered by incorrect
meter reads due to meter reading errors and not leaks. This large exceptions rate appears to have a
direct correlation with the high levels of adjustments discussed in the previous section.
Draft Final TM3-7

Meter reading error rates appear to be consistent with both standard and AMI meters, thereby
eliminating AMI as a primary driver for the high exception rates overall. There are currently no quality
assurance standards deployed during the meter reading process (although the current handheld system
has some built in functionality that, if used, could help). As a result, bad meter reading data is uploaded
into the billing system and putting tremendous pressure on staff to catch large variances prior to bill
mail out. There is an existing manual process by which daily threshold exception logs are reviewed by 3
or 4 edit clerks. Once identified, the high bills are pulled prior to mail out and reviewed. The bill is then
adjusted and estimated bills are sent out. Because of the manual process and the high volume of
exceptions, it is unlikely that most of the exceptions are addressed prior to mail out. A large percentage
of customer bills are rendered incorrectly and require adjustment.
Concurrent with the Project Team Review, the City also obtained separate information from the AMI
meter installation contractor suggesting that a number of meters were using gallon units. The majority
of these identified meters were installed after the time of the Project Team data extract. Those that
om
were installed prior to the extract had not registered unusually high consumption.
.c
Misapplication of Billing Policy
Minimum Sewer Charge
ya
The most critical finding by the Project Team is the incorrect calculation of the minimum sewer charge.
The City’s Code of Ordinances, Division 2, Sec. 122-235 regarding minimum monthly sewer charges
la
states:
ba
(a) The schedule of minimum sewer service charges is as follows:

(1) Each customer with a five-eighths-inch meter whose water consumption is 300 cubic feet or less
am
shall be assessed a minimum monthly charge of $13.41.

(2) Each customer with a one-inch meter whose water consumption is 670 cubic feet or less shall be
nj
assessed a minimum monthly charge of $29.95.

so
(3) Each customer with a one and one-half or two-inch meter whose water consumption is 1,510
cubic feet or less shall be assessed a minimum monthly charge of $67.50.
ck
(4) Each customer with a three-inch or larger meter whose water consumption is 2,710 cubic feet
or less shall be assessed a minimum monthly charge of $121.14.
Ja
Customers are billed bimonthly, and the monthly minimum sewer rate as described in the ordinance is
in fact charged only bimonthly, meaning that all users below the minimum threshold are charged only
half the expected amount. This lost revenue equates to at least the following levels of lost revenue
between Nov 8, 2013 and Oct 17, 2014.
Table 3: Lost Revenue from Billing Misapplication
METER SIZE LOST REVENUE

5/8” $ 630,394
1” $ 100,546
1 1/2” AND 2” $ 150,052
>2” $ 35,615
Draft Final TM3-8

TOTAL $ 916,607
In Figure 2, below, the number of 5/8” meters charged $13.41 and the number of times each account
received this charge between Nov 8, 2013 and Oct 15, 2014 (nearly a full year after the rate increase) is
plotted. This amount should never be billed, but it happens thousands of times. Rather, twice the
monthly charge, or $26.82 for 5/8” meters, should be minimum charge found on a bill.
7,000
6,000
5,000
Number of Meters
4,000
om
3,000
2,000
.c
1,000
ya
1 2 3
la 4 5 6
Number of $13.41 sewer charges over 6 billing cycles of 2014
ba
Figure 2. Instances of Half Minimum Sewer Charge for 5/8" Meter

am
nj
Hung Accounts
City staff also mentioned that the billing system was set up in such a way that accounts could be set to
so
non-bill status even though water service may still be available. Hung accounts are those that that are
listed as active but have multiple instances of consumption greater than zero, and no charge for water
ck
and sewer. The Project Team identified 73 accounts in the legacy billing system where this happened
Ja
continuously between January 2013 and October 2014. These accounts represented 1,374,318 cubic
feet of unbilled consumption over this period, which based on current rates would be about $45,000.
Upon further review during the field audit, the consultant team found most hung accounts to be
resolved. As of the date of the audit, only 20% of accounts on the sample list were still unresolved.
According to Information Systems staff, intensified efforts have been recently placed on resolving hung
accounts as part of the data cleanup for conversion.
Vacant Properties with Registered Consumption

The Project Team noticed when reviewing the billing data in bulk that there are many instances of
incomplete or inaccurate customer information, which could lead to unenforceable collections
circumstances. Specifically, there are accounts where the customer is “No Present Occupant,” not
receiving a bill, but with consumption registering regularly. The Project Team generated a list of 1,304
Draft Final TM3-9

“No Present Occupant” accounts representing 2,850,201 cubic feet of unbilled consumption during the
last two-month period of the legacy billing system. The time constraint on this is meant to avoid any
issues related to a change in customer status over time, since the customer information is just a
snapshot as of the time the Project Team obtained a copy of the billing system.
The team identified a sample of 40 accounts and received additional information on 38 of the 40
requested samples, of which only 23 (58%) were still in NPO status at the time of the field audit.
Seventeen (74%) had been closed out as part of the standard NPO process, while 6 (26%) hadn’t been
closed out. Twenty-one (91%) of the 23 accounts had outstanding balances, which means the vacating
customer didn’t settle their final bill.
Non-Billed Accounts
There are several types of non-bill status in which an account can fall. They are:
om
 Off- Customer Request
 Customer moving (off)
 Off- Delinquent Bill
.c
 Off- Bad Check
ya
 Off- Seasonal
 Returned mail
la
 Special Payment Plan
ba
These categories relate to circumstances in which a bill should not be generated because no
am
consumption is allowed. However, the Project Team found 410 instances of non-bill accounts with
consumption activity between September 1, 2014 and October 15, 2014. This activity totaled 769,734
cubic feet of consumption or $53,417.
nj
so
The field audit included a sample of 43 accounts, from which the consultant team identified a lack of
follow-up on the cause of non-bill status to be the main driver. After an account has been classified as
ck
non-bill status, it appears that accounts go unread for upwards of a year after last billed date. It also
appears that field actions to pull meter or tie-in also take a considerably long time. So, if an account is
Ja
shifted to “Off- Bad Check,” for example, it could be a long time before a field technician actually shuts
off the meter (if ever). Secondarily, the process for moving accounts to non-bill status is being exploited
by customers. Customers that are delinquent on their payment get switched to a non-bill status and so
stop getting billed. Since field work orders to shut off account are not closely monitored to confirm that
actual shut off happened, customers continue to get free service for extended periods of time.
Although, meter reads continue to confirm usage on these accounts subsequently, actions to bill or
terminate service physically, still doesn’t happen for extended periods of time. Even when physical shut
off eventually happens, we were made to understand that new service for the location can be resumed
by just using another relative’s name. This system is fraught with opportunities for service usage without
a bill.
Draft Final TM3-10

Outstanding Balances
As shown in the high level data review, collections have decreased over time, with a recent steep drop.
Concurrently, outstanding balances have been steadily increasing. The Project Team identified 17,214
accounts with outstanding balances as of the most recent bill occurring in 2014. This totals $15,657,909,
or about $900 per account. The field audit of a sample of these accounts resulted in further
acknowledgement of the process lapse between shut off work order request and actual enforcement
action.
Accounts flagged with the government or extension (“G” or “E”) code on their meter to prevent shutoff
were also of particular interest as they make up 2-3% of accounts, but account for 17% of outstanding
balances as of the end of available data.
Exempt Government Accounts

Accounts belonging to the local, State, and Federal government are coded as “G” accounts and not
om
subject to shut off. At the time of the bulk data review, there were 575 “G” coded accounts with an
outstanding $ 1,070,856 on their most recent bill. State properties in particular had an outstanding
.c
balance of $10,142 just on bills generated in September and October of 2014 (presumably those
accounts still actively in use that would feel the effects of enforcement action).
ya
During the field audit, the consultant team confirmed that most (92%) of the sampled accounts were
la
either state or local governmental facilities. However, we also identified 2 (8%) non-profit organizations
ba
on the sample list. These 2% should have been coded “E” accounts, but since the result for the two is
similar, this was not concerning. The team was able to confirm all accounts were metered and billed
am
routinely. However, there does not appear to be any robust or active collection efforts on these
accounts, allowing nonpayment to perpetuate without consequence.
nj
Exempt Extension Accounts

so
An account code of “E” is intended for those customers for whom the City cannot turn off water service.
They represent hospitals, sick or elderly residents, those customers on payment plans, and other subsets
ck
of the population for which it may be irresponsible to shut off water service.
Ja
During the field audit, the team reviewed 37 accounts and found various discrepancies. While it appears
this status was originally setup for accounts classified as emergency facilities, due to system limitations
it is now being used primarily for payment plan to exempt accounts from shut off and for customers in
poor health.
Utility documentation suggests that this code should only be entered by a supervisor, however, no
official policy nor appropriate authorization procedures for classifying customers as “E” or setting up a
payment plan was provided. The “E” coding screen has no control; any user can go into the system and
add an “E” code without approval. Generally, the controls guiding the processing of these transactions
are very weak; it is not clear how customers with health issues are identified, classified and authorized.
This categorization represents a clear opportunity for customers to essentially sign up for exempt status.
Draft Final TM3-11

Over 60 percent of customer account sample reviewed were not complying with current payment plan
and no action has been taken to escalate or take them off this “E” status.
In recent months, Information Systems staff has begun generating a monthly report of “E” customers on
payment plans (information about which is included in a comments field) for review by Customer
Service management. This is intended to ensure that payment plans are being honored and to allow for
enforcement action where this isn’t the case. However, given the slow turn-off process, this is likely not
to result in significant improvements in collection.
Aged AR Balances
The team identified a sample of 40 accounts and received additional information on 37 (93%) of the 40
requested samples. Based on review of this information, 23 (62%) of the accounts had been shut off in a
timely manner for nonpayment. As of the date of the audit, 8 (35%) of the accounts appeared to still be
using water. This exemplified the team’s concern that the utility does not have any active or aggressive
om
collections process. The current process requires delinquent accounts to be forwarded to a third party
collection agency. However, we discovered that no accounts had been sent for collection in almost 12
.c
months.
ya
The results of the bulk billing data review and categories of noted exceptions are summarized in Table
below.
la
ba
Table 4. Summarized Exceptions found during Bulk Analysis

am
CATEGORY DESCRIPTION TIME NO OF CONS (CF) DOLLAR VALUE

PERIOD INSTANCES
nj
ADJUSTMENTS – Positive or negative 1/1/13 - 3,391 $ (41,204,999)

LARGE adjustments exceeding or 11/7/14
so
equal to $500
ADJUSTMENTS - Accounts with more than 1/1/13 - 141 $ (2,562,740)
ck
FREQUENT 4 adjustments during the 11/7/14

period
Ja
HUNG Accounts registering 1/1/13 – 73 1,374,318

ACCOUNTS consumption with no 10/15/14
water or sewer charges
STUCK AMI Meters with steady 1/1/14 – 283
METERS consumption prior to AMI 10/15/14
meter installation that
drops to or near zero after
installation
AMI METERS Meters with readings 1/1/14 – 3,371
ABOVE greater than 2 standard 10/15/14
THRESHOLD deviations above the
mean of reading prior to
AMI meter change
Draft Final TM3-12

“NO PRESENT Accounts with no present 9/1/14- 410 769,734 $ 53,417
OCCUPANTS” occupant at time of 10/15/14
WITH METER extract that are
READINGS registering consumption
OTHER Accounts with non-bill 9/1/14- 1,304 2,850,201
ACCOUNTS IN status at time of extract 10/15/14
NON-BILL that are registering
STATUS consumption
“G” ACCOUNTS Accounts with meters As of 575 $ 1,070,856
with a government code 10/15/14
of ‘G’ at time of extract,
sum of outstanding
balance
“E” ACCOUNTS Accounts with meters As of 1,140 $ 1,613,439
with a government code 10/15/14
om
of ‘E’ at time of extract,
sum of outstanding
balance
.c
AGED Total outstanding balance As of 17,214 $ 15,657,909
ya
ACCOUNTS as of 11/7/14 for all 11/7/14
RECEIVABLES accounts that received at la
least one bill in 2014
ba
am
nj
so
ck
Ja
Draft Final TM3-13

Potential Fraudulent Activity
Fraudulent activity may be partially to blame for the lower than expected revenues. The Project Team
considered several possible types of fraud that could be committed, and determined which would be
identifiable from analysis of the billing data. These included:
 Unwarranted adjustments: This could be an adjustment performed for family or friends in which
the customer is complicit or an adjustment in lieu of a registered payment (either of which
would lower an account balance) when the customer pays at the cashiering counter.
 Intentional service order to stop billing: Because service orders entered in the system suspend
billing while the order is being fulfilled, this could be a method of keeping a customer’s service
turned on despite nonpayment. Often, service orders are resolved slowly, so the non-billed
status could be in place for quite a while.
 Non-bill status: There is a code within the legacy billing system that denotes when no bill should
be generated for a particular account. Typically a non-bill status is triggered by a work order (as
om
described above) but this field can also be altered manually to achieve the same result.
 Falsified meter reads: Meter reads that are intentionally entered incorrectly would result in
.c
lower bills for some customers (at least until a true up occurred). This would appear the same in
ya
the system as a true read and is difficult to tease out, especially where it has been happening for
the same meters over an extended period of time such that consumption appears to be lower
la
than normal, but typical for that customer.
 Misapplied exemption flag: Another flag in the legacy billing system that prevents enforcement
ba
action is the G/E flag. Customers considered to have “G” (government) or “E” (emergency)
accounts cannot be shut off for nonpayment. Despite the intent, this code could be used to
am
prevent shut off for accounts that do not qualify for “G” or “E” status. There is no audit history
on this code included in the daily files.
nj
 Falsified “E” status: Accounts can be switched from Active statuses (coded 1, 3, or 9) to an
“Extension Customer” status of E, which extends the date by which a payment is due and keeps
so
nonpayment from triggering shutoff procedures.

ck
 Registered payments that are not deposited to the bank: With proper accounting oversight and
settlement procedures, this should not be a problem. However, the available data only shows
Ja
what payments were registered, not how much money is deposited to the bank. This
opportunity for fraud is included for completeness sake but is unlikely to be a true concern.
 Field tampering/meter bypass, disconnect bribes: These items were suggested by City staff, but
cannot be identified within the available billing data.
Because of some limitations in the legacy billing system auditing and the nature of how the billing
system receives information from meter readers, there are several categories of potential fraudulent
activity that are indiscernible through available data. However, three primary categories—adjustments,
service orders, and “E” statuses—can be investigated through the use of a nightly user journal and a
nightly work order history. Neither the adjustments data nor the service order data show suspicious
transactions.
Draft Final TM3-14

However, the project team found evidence suggesting fraudulent activity in the use of “E” status codes.
The team reviewed transactions where an account was coded “E” and compared the application user
making the change with the benefiting accountholder. Details of this review have been submitted to the
Director of Public Works for further review.
Outside 1 Mile Perimeter

Customers more than a mile outside the City (herein called the “Well” service area) are subject to a
different set of rates than those within a mile of the City boundary or within the City itself (herein called
the “Plant” service area). Well customers only receive sewer service from the City, and sewer rates in
that service area have to be approved by the Mississippi Public Service Commission. Historically, these
rates have been 33.636% of Plant area rates. The City approved rate increases for Plant area customers
subject to its primary rate schedule on January 1, 2011 and then another one on November 13, 2013.
The City may not have gotten approval from the PSC for one or both of those rate increases to extend to
the Well service area. As such, these customers may have been overcharged by the City for a period of
om
time (one or nearly three years) and the City is interested in refunding any bills issued and paid in error.
The Project Team has identified these customers within the legacy billing system and compiled their bill
.c
history.
ya
Next Steps
To confirm the total estimated overcharged amount, the Project Team must better understand during
la
which rate increase approval by the PSC was either not sought or not received. At that point, charges
ba
levied to those customers during the legacy billing system’s two-year history can be compared with
correct charges as calculated with the most recent approved Well area rate to determine the difference.
am
The Project Team will estimate overcharges occurring prior to the earliest billing history in the legacy
system based on daily histories, though those have less account detail than the legacy system.
nj
After we determine the total estimated overcharged amount, we can consider matching up payments
so
and adjustments to bills to determine whether the difference between the billed amount and the
correctly calculated charge needs to be adjusted to accommodate either nonpayment or an adjustment
ck
already recorded. This effort will be undertaken depending on the available resources and time.
Ja
RECOMMENDATIONS
Business Processes
Auditing
Billing system users have a great deal of authority to edit account and customer characteristics without
much oversight. While staff discussed the oversight and approval processes, the data told a somewhat
different story. Some individuals suggested that user IDs and passwords are known and used by multiple
staff members. If an approver’s credentials are not private, the authority those credentials carry can be
abused as well. The Project Team recommends that the City develop complete standard operating
procedures for how and when to make certain kinds of edits in addition to a thorough edit review and
approval process, based on bulk review of a user’s transactions over a certain period of time. There
should be thresholds for adjustments above which additional approval is necessary.
Draft Final TM3-15

If this cannot be accommodated in the legacy system, an alternate method should be used in the interim
until the conversion to CC&B is complete.
Work Orders
The most considerable breakdown in customer service processes seems to be in the work order system.
When a work order is entered, that account is switched to a non-bill status and the work order is sent to
the team of field technicians. Once that automated process is complete, there is no requirement that
the field technician log completed work in the system, so work goes undone for great lengths of time.
During this time, consumption continues without being billed. The field technician operation was
described as a mostly unsupervised and often paper-based, resulting in poor performance. This is
corroborated by the billing data and suggests that the City should take immediate action to improve the
work order process.
In the short term, this should include hiring capable managers to oversee staff work, check that work
om
orders are completed in a timely manner, and verify that work is being done thoroughly. In the longer
term, there should be a mechanism for closing the loop on work orders within the system that
.c
generates and manages them. This is an important consideration for the conversion process as well; the
City needs to ensure that the current process is not replicated in CC&B but improved instead.
ya
Conversion
la
As of early 2015, the utility had suspended its efforts to convert to Oracle’s CC&B customer information
ba
and billing software. However, should this proceed in the future, there will be a great opportunity to
establish measures to improve data quality so that billing going forward is more accurate and more
am
complete.
nj
Data Flags
Within the legacy billing system, there are several circumstances that are signified by a combination of
so
other flags. For example, the summer sewer discount of 15% is given to accounts that are categorized
separately as “Residential” and “House,” and as including sewer service. The correct application of this
ck
discount relies on data in three fields being accurate. While the hope is to have all data be as accurate as
possible, opportunities for the discount being applied or not applied incorrectly are increased when
Ja
there are three times as many potential data errors. Similarly, there are instances where charges were
not replicable and no flag was found in the billing system to suggest whether another factor was at play
(an undocumented discount). Similarly, in some cases the same field being used to signify more than
one circumstance, such as the “E” field signifying both healthcare facilities and individual customers
exempt from meter shut off). The new billing system should allow for further categorization of accounts,
meters, and customers, such that derivation of a particular category or combination of categories is no
longer necessary.
Auditing
The new billing system can also be set up with more robust auditing processes. This would include both
unique and private user IDs under which to perform customer service transactions as well as more
streamlined and thorough supervisory approval processes.
Draft Final TM3-16

Data Validity
Going forward, there should be data validity checks performed on all entries before they are committed
to the system. In the legacy system, there are invalid entries in nearly every field and table, as described
above. This could result in in accurately calculated bills if the calculation relies on a defined set of valid
values. CC&B should bet set up not to accept invalid values in any field.
Warehousing
The nightly reports provided a tremendous amount of historical information for use in this analysis.
However, there is a great deal of room for them to be more informative. A data warehousing process
should be considered during the transition to capture frequent, more complete copies of the database.
Enforcement
The current billing system allows for delinquency or other conditions to trigger a work order, but there is
om
no follow-through on whether the work order is ever satisfied. In many cases, the work does not happen
in a timely fashion and customers are allowed to continue receiving water service despite nonpayment.
.c
The new system should close the loop on work order generation and completion, and customers should
continue to be billed for all consumption registered, whether or not their account is delinquent.
ya
During conversion, it is likely that the outstanding balance will be converted but other information about
la
the reason behind the balance or the age of receivables may not be. The City should carefully consider
ba
its enforcement policies at the time of conversion to avoid instantly shutting off customers who have
been receiving service without payment for long periods of time. Rather, the City should spend
am
additional effort to communicate the terms of enforcement action and collect on those accounts before
shutting them off.
nj
Many of these concepts have likely been included in the design of CC&B, but as they represent a critical
so
shift in data structure, policies, and procedures, they remain important to keep in mind as data are
transferred from the legacy billing system to CC&B.
ck
Ja
Draft Final TM3-17

Appendix A. Billing Data Field Audit Report
Conducted by Intel Business Solutions
SCOPE
Given the significant decline in incoming water and sewer revenue, the objective of this task is to acquire
utility billing system data from the City and review it to identify patterns in billing and collections that may
explain recent reductions in collections. The collection rate fell significantly between 2012 and 2013,
current revenues are materially below expected totals, and as indicated in our recent meeting, we will
assist in determining if as many as a quarter of the accounts are not being billed in any given billing cycle.
In other words, the Project Team will answer the following questions:
om
 Who is affected?
 How it is occurring?
 Why it is occurring?
.c
 How much revenue is potentially at stake?
OBJECTIVES
ya
la
ba
Overall, this task will accomplish the following:
 Identification of unbilled accounts and a determination whether there are one or two dominant
am
problems (reasons for an account not being billed) or a number of small problems;
 Assuming there is a dominant problem, whether patterns are identifiable;
nj
 Specific information on a small set of customers impacted by the dominant problem;

 A range of additional revenue that may be generated from resolving the problem; and
so
 Development of a package of information summarizing these findings to take to the Mayor.

ck
In order to accomplish these objectives we focused on the following three audit assertions:
Ja
1. Completeness – ensure within reason that all accounts are billed regularly and timely; also that
all balances that should have been recorded are recorded.
2. Existence – ensure within reason that all accounts billed actually exist and all balances carried also
exist.
3. Accuracy and valuation – ensure within reason the accuracy of billings and value of account
balances carried.
Transactional/balance Areas of Focus
1. Metering/Consumption data
2. Billings/Revenues
3. Collections/Accounts Receivable
Draft Final TM3-A1

TECHNICAL APPROACH/WORK PLAN
1. Fact gathering and understanding

a. Meet with city staff including, IT, billing department and others
b. Document request including, policies, procedures, system requirements, data dumps etc.
c. Follow-up meetings with relevant staff
2. Research and data diagnostic

a. Analyze data
b. Identify gaps and trends
3. Field audit
a. Undertake field research/audit on specific issues and accounts. See audit objectives in the
om
field audit section.
.c
4. Prelim findings
a. Develop findings including estimates of impacts
ya
b. Develop recommendations
c. Discuss draft findings and recommendations with client
la
ba
5. Final findings and recommendations
a. Finalize report and issue
am
STATUS UPDATE
nj
Data Diagnostics
so
Preliminary data diagnostics results based on analysis of data from 1999 to 2014 (October) revealed:
ck
1. Gross Consumption is consistent with prior year trends of gradual decline –

Ja
completeness/valuation and accuracy

2. Number of accounts billed also appear to be consistent with historic trends of gradual decline –
completeness
3. Rates schedule appears to be correctly applied for most accounts, with some exceptions noted –
valuation and accuracy
4. Large number and size of adjustments noted in 2014, which continued a trend of increasing
annual adjustments – valuation and accuracy
5. Collections rate showed a continued trend of decrease, with significant decreases in 2013 and
2014
The following exceptions were noted for further field review and follow-up:
Draft Final TM3-A2

1. Minimum sewer rate appears to be incorrectly applied – ordinance requires a monthly minimum,
but many accounts appear to be billed bi-monthly.
2. Other exceptions noted included:
a. Adjustments – large
b. Adjustments - frequent
c. Hung Accounts
d. Stuck AMI meters
e. AMI meters above threshold
f. Accounts Not being billed - flagged as “No Present Occupants” with meter readings
g. Other accounts in Non Bill Status
h. Accounts coded as “G” and not subject to shut off
i. Accounts coded as “E” and not subject to shut off
j. Aged accounts receivables
om
Field Audit
.c
Field work commenced on December 8th with the following activities:
ya
1. Developed sample data for further investigation
2. Met with billing department and IT staff to determine best method or approach to access data
la
required
ba
3. Met with billing and IT staff to further understand and document billing practices and adjustment
procedures
am
4. Requested and received additional data for further analysis of AR balances

5. Reviewed some of the additional data provided and discovered a large number of accounts in the
nj
system are coded in Non Bill status but continue to record consumption - appears that more data
diagnostic is needed for completeness
so
a. This exception group was then added to the audit list.

ck
FIELD AUDITS
Ja
Field visits were undertaken in both December (18th to 19th) and January (7th to 10th) with review of the
following exceptions:
1. “G” Accounts – these are government G Accounts Count %

accounts exempted from shut off. We
Sample Size 24 100%
reviewed 24 accounts with results
Government Accounts 21 88%
summarized in table.
NPO's 3 13%
a. Compliance with policy/practice – most (88%) of the sampled accounts were either state or
local governmental facilities, however, we also identified 3 (13%) non-profit organizations on
the sample list and 3 (13%) No Present Occupant accounts.
b. Metering and Billing – we were able to confirm all accounts were metered and billed routinely.
Draft Final TM3-A3

c. Collections – there does not appear to be any robust or active collection efforts on these
accounts.
2. “E” Accounts – These are accounts that are either E Accounts Count %
emergency type facilities or accounts on payment Sample Size 37 100%
plans or individuals classified with health issues. Did Not Receive Information. 4 11%
Received Information 33 89%
This code exempts accounts from being cut off. % of Individuals 32 86%
We reviewed 37 accounts with results Summary of Collected Information
summarized in table: Count %
Payment Plan 28 85%
No Payment Plan 5 15%
a. We received backup information on 33 of Analytics
%Count
the 37 requested, of which ~85 percent
Remove E Code; No Current On PP 54% 15
of customer has payment plans. In had Do Not Remove E Code; Current on PP 46% 13
payment plans were not current on their
om
payment plan and no action has been taken to escalate or take them off this e status.
.c
i. Only 28 (85%) were on a payment plan as required to be coded in this status, of which:
1. 15 (54%) were not adhering to the payment terms and should have been
ya
removed from the “E” code; and
2. 13 (46%) were making regular payments in line with the payment plan.
la
ii. 5 (15%) of the accounts reviewed were not on a payment plan and did not appear to
ba
be appropriately coded as “E” account.

am
b. Other issues noted include:

i. Appears this status was originally setup for accounts classified as emergency facilities,
nj
however due to system limitation it is now being used primarily for payment plan to
exempt accounts from shut off and for health related customers
so
ii. No official policy nor appropriate authorization procedures for classifying customers
as “E” or setting up payment plan was provided
ck
iii. The controls guiding the processing of these transactions are very weak and therefore
prone to abuse and/or fraud.
Ja
1. Not clear how customers with health issues are identified, classified and
authorized
2. Essentially any customer can request for this and any customer service rep
can execute transaction without any oversight or authorization
3. The E coding screen has no control. Anyone can go into the system and add
an E-Code without any oversight
4. The Field alteration screen has the capability of changing nearly everything
on a customer. The only thing that cannot be changed is the customer ID and
the Premise location.
Draft Final TM3-A4

3. Stuck AMI Meters – These are AMI meters that are Stuck AMI Meters Count %
not registering readings properly and need to be Sample Size 20 100%
Did Not Receive Information. 0 0%
replaced or fixed. We pulled a sample of 20
accounts with results summarized in table: Summary of Collected Information
Count %
Resolved Problem 18 90%
a. Timeliness of resolution: Problem Not Resolved 10%2
i. According to client the standard Analytics
for resolution is 8 weeks (no Count
%
Hand No Moving 17%3
policy or SLA was provided). Our Meter Was Off 28%5
review revealed that about 90% Installed New Meter 33%6
Located Meter To Read 17%3
meet this standard - based on
Ticket Not Returned 6%1
service order notes in the sample
group. As of date of the audit, only 2 accounts (10%) on the sample list were still
unresolved.
om
1. However, service order process could be improved – doesn’t currently appear
to be a process for closing out service orders with the billing system. Lots of
.c
opportunities for service orders to fall between the cracks.
ii. Customer billing – unbilled periods are estimated using the most recent month’s bill
ya
b. Based on our review of the work order tickets we were unable to determine whether the root
la
cause of the issue was meter defect or improper meter installation – not enough detail is
ba
provided on the job ticket.

am
4. Hung Meters – These are non AMI meters that are

Hung Meters Count %
not registering readings properly and need to be
Sample Size 10 100%
nj
replaced or fixed. We pulled a sample of 10 Did Not Receive Information. 0 0%

accounts with results summarized in table: Received Information 10 100%
so
Summary of Collected Information

a. Timeliness of resolution: Count
%
ck
Resolved Problem 80% 8

i. Meter issue – according to client
Problem Not Resolved 20% 2
the standard for resolution is 8
Ja
Analytics
weeks (no policy or SLA was Count
%
provided). Our review revealed Different Meter 13% 1
that about 80% meet this standard Meter Was Off for Readouts 50% 4
Installed New Meter 25% 2
- based on service order notes in
Read Meter 13% 1
the sample group. As of date of
the audit, only 2 accounts (20%) on the sample list were still unresolved.
1. Service order process could be improved – doesn’t currently appear to be a
process for follow-up on open service orders.
ii. Customer billing – unbilled periods are estimated using the most recent month’s bill.
b. Based on our review of the work order tickets we were unable to determine whether the root
cause of the issue was meter defect or improper meter installation – not enough detail is
Draft Final TM3-A5

provided on the job ticket. Anecdotally, we believe this might be symptoms of aging/old
meters.
5. No Bill Status – These are accounts that have been

classified as Non Bill Status but continue to
No Bill Status Samples Count %
Sample Size 40 100%
register meter reads. We pulled a sample of 40
accounts with results summarized in table below: Received Information 39 98%
Analytics
a. Based on our review of the 39 samples Count %
and work order tickets received it appears Have a Service Order 39 100%
that accounts continue to register Pulled Meter 24 62%
readings for upwards of a year after last Pulled Tie-In 8 21%
billed date (date reclassified to No Bill Inaccessible Meter 2 5%
Status). While it appears that field actions Meter On 1 3%
om
Meter Turned Off 2 5%
are not always timely, it was clear from
Faulty Meter (Hand Not Moving) 1 3%
the tickets that repeated field actions Meter Not Pulled 1 3%
.c
were carried out on many of the accounts
during the one year period of continued water consumption after initial shutoff notice. These
ya
field actions were reactions to customer theft issues, as customers appear to routinely turn
meter back on after shut off, resulting in the City ultimately pulling the meter and or tie in.
la
ba
6. NPO – these are accounts classified as No Present NPO Samples Count %

Occupant (non-bill status) and continue to register
Sample Size 40 100%
am
meter reads. We pulled a sample of 40 accounts

with results summarized in table: Received Information 38 95%
nj
Existed as NPO's 23 58%

a. We received backup information on 38 of
so
Summary of Collected Information

the 40 requested samples, of which only Count %
23 (58%) were still in NPO status at the
ck
Not Closed Out 6 26%

time of the field audit. Closed Out 17 74%
Ja
AR Balances 21 91%
i. 17 (74%) had been closed out as
part of the standard NPO process, while 6 (26%) hadn’t been closed out.
ii. 21 (91%) of the 23 accounts had outstanding AR balances, which means the vacating
customer didn’t settle their final bill.
Draft Final TM3-A6

7. Meter Reading threshold – These are meters that
are registering readings that are significantly
Meter Reading Threshold Count %
higher the average consumption. We pulled a
Sample Size 30 100%
sample of 30 accounts with results summarized in Did Not Receive Information. 0 0%
table below: Received Information 30 100%
Analytics
a. DPW currently experiences extremely high Count %
threshold exception rate, averaging Faulty Meters 7 23%
approximately 53% of all billed accounts Misreads 21 70%
needing review. Accurate - leaks 2 7%
b. 21 (70%) of the exceptions appear to be triggered by incorrect meter reads due to meter
reading errors and not leaks. Also, this large exceptions rate appears to have a direct
correlation with the high levels of adjustments discussed in the next section.
om
i. Meter reading error rates appear to be consistent both pre AMI and post AMI thereby
eliminating AMI as primary driver for the high exception rates. There are currently no
.c
QA standards deployed during the meter reading process (although, current
handheld system has some built in functionality that if used could help) resulting in
ya
bad meter reading data being uploaded into the billing system and putting
tremendous pressure on billing/edit clerks to catch large variances prior to bill mail
la
out. Secondarily, because of the manual process and the high volume of exceptions,
ba
it is unlikely that most of the exceptions are addressed prior to mail out. Therefore a
large percentage of customer bills are rendered incorrectly.
am
ii. Current manual process – daily threshold exception logs are reviewed by 3 or 4 edit
clerks. Once identified the bills are pulled prior to mail out and reviewed. The bill is
nj
then adjusted and estimated bills are sent out

so
8. Many Adjustments – These are accounts that

have a high frequency of adjustments. We pulled
Many Adjustments Count % Amount %
ck
a sample of 159 accounts with results Sample Size 159 100% ($2,359,164) 100%
summarized in table: Did Not Receive Information. 107 67% ($1,347,442) 57%
Ja
Received Information 52 33% ($1,011,722) 43%

a. We received backup information on 52 Summary On Collected Information
(33%) of the 159 requested samples. Count % Amount %
Meter Leaks 8 15% ($24,570) 2%
i. 44 (85%) of the work orders
Meter Misreads 44 85% ($987,152) 98%
revealed issue to be caused by
Total 52 100% ($1,011,722) 100%
meter reading errors
ii. 8 (15%) of the work orders were due to meter leaks
b. Policy and controls – there is a leak adjustment policy and a standard process for meter error
adjustments. All the adjustments sampled were processed in accordance with process
Draft Final TM3-A7

described and were properly authorized, however, there does not appear to be a signoff
threshold.
i. Every adjustment has a preparer and an approver
9. Large Adjustments – These are accounts that have

Large Adjustments Count % Amount %
large adjustments. We pulled a sample of 61
Sample Size 61 100% ($2,376,101) 100%
accounts with results summarized in table: Did Not Receive Information. 31 51% ($1,008,225) 42%
Received Information 30 49% ($1,367,876) 58%
a. We received backup information on 30 Summary On Collected Information
(49%) of the 61 requested samples. Count % Amount %
i. 16 (60%) of the work orders Meter Leaks 2 7% ($12,169) 0.9%
Meter Miss Reads 18 60% ($1,246,821) 91.2%
revealed issue to be caused by
Faulty Meter 2 7% ($145,964) 10.7%
meter reading errors – miss reads Other 8 27% $37,078 (2.7%)
ii. Only 2 (7%) of the work orders
om
Total 30 100% ($1,367,876) 100.0%
were due to meter leaks
.c
10. AR – These are accounts that have aged
ya
AR balances. We pulled a sample of 40
AR Accounts Count %
Sample Size 40 100%
accounts with results summarized in
la
Not On An AR Account 1 3%
table below:
ba
Did Not Receive Information 2 5%
a. We received backup information
am
Analytics
on 37 (93%) of the 40 requested
Count %
samples. Based on review of Initial shut off timely 23 62%
nj
backup: Initial did not shut off timely 14 38%

Shut Off Analytics
so
i. 23 (62%) of the accounts Count %

had been timely initially Account Still Reading Consumption 8 35%
ck
shut off for nonpayment. Still Off From Initial Shut Off 15 65%
As of the date of the
Ja
Did not shut off Timely; Still On 5 36%

audit, 8 (35%) of the
accounts appeared to still be using water.
ii. 14 (38%) of the accounts had not been timely shut off
b. The biggest issue we discovered was that the utility does not have any active or aggressive
collections process. The current process requires delinquent accounts to be forwarded to a
third party collection agency. However, we discovered that no accounts had been sent for
collection in almost 12 months.
Summary Results
Based on our efforts to date, we have grouped the issues identified within three classifications:
1. Unbilled Revenues – many accounts continue to receive services and are not timely or routinely
billed, primarily due to untimely meter shut offs and significant water theft issue.
Draft Final TM3-A8

2. Ghost Revenues – many accounts fail threshold exception edits primarily due to meter reading
errors resulting in large and frequent adjustments.
3. Uncollected Revenues – many accounts are delinquent and have significant accounts receivable
balances. However the city does not have an active/proactive collections process.
om
.c
ya
la
ba
am
nj
so
ck
Ja
Draft Final TM3-A9

Analysis of the Performance Guarantee Component of the Siemens Contract
Raftelis Financial Consultants, Inc. (RFC) was engaged by the City of Jackson (City) to review and analyze
the existing contract between the City and Siemens Industry Inc. (Siemens). The purpose of the review
and analysis was to develop recommendations for actions the City could take to maximize the value of
the Siemens contract.
OVERVIEW
In January 2013, the City entered into a contract with Siemens to make improvements to the City’s
water and sewer systems. Siemen’s scope under the contract consisted of three major tasks: 1)
Advance Metering Infrastructure (AMI) Upgrade; 2) Water Treatment Plant (WTP) Repairs and
Upgrades; and 3) Sewer Collection System (SCS) Repairs. The total fixed cost of these projects to the
om
City is $90,983,106 with additional payments due to Siemens for the Performance Assurance Program
services.
.c
The WTP Repairs include a variety of projects at the JH Fewell and OB Curtis WTPs. The total cost of
ya
these projects is $10,969,673 with the majority of the cost associated with projects at the OB Curtis
la
WTP. The SCS Repairs include approximately 20 projects involving repair, replacement and/or
relocation of sewer collection system infrastructure. The total cost for the SCS Repairs is $15,844,194.
ba
The AMI Upgrade is the largest component of the contract and consists of the installation of almost
am
65,000 remote read water meters; the infrastructure necessary to collect data from the remote read
meters; and a new customer billing system. The total cost of the AMS Upgrade is $51,209,884 or 56% of
nj
the contract value. An additional $12,959,355 in contract costs is related to Development, Parent
Company Guarantee and Mobilization. Table 1 below summarizes the contract tasks and the cost for
so
each.
ck
Table 1 – Siemens Contract Tasks and Costs

Ja
Contract Task Cost

Development $ 1,120,000
Parent Company Guarantee $ 148,000
Mobilization $ 11,691,355
Billing Software $ 11,320,444
Water Meters $ 39,889,440
JH Fewell WTP $ 3,980,465
OB Curtis WTP $ 6,989,208
Sewer lines $ 15,844,194
Total Cost $ 90,983,106
Draft Final TM4-1

The anticipated benefit of the AMI Upgrade project will be improved revenue collection resulting from
more accurate water meter data as well as other operational savings and savings accruing as a result of
deferred maintenance. Table 2 below is a reproduction of Table 1.2 from Exhibit C of the Siemens
contract and shows the total savings that Siemens guarantees will accrue to the City as a result of the
contract.
Table 2 – Guaranteed Savings

Large
Small Meter Meter
Billable Billable Deferred
Performance Usage Usage Operational Maintenance
Period Increases Increases Savings Savings Total Savings
Construction $484,347 $501,802 $503,750 - $1,489,899
om
Annual Period 1 $2,421,737 $1,003,604 $2,015,200 $1,750,000 $7,190,541
Annual Period 2 $2,555,055 $1,003,604 $2,075,656 $1,750,000 $7,384,315
Annual Period 3 $2,688,373 $1,003,604 $2,137,926 $1,750,000 $7,579,903
.c
Annual Period 4 $2,821,691 $1,003,604 $2,202,063 $1,750,000 $7,777,358
ya
Annual Period 5 $2,955,010 $1,003,604 $2,268,125 $1,750,000 $7,976,739
Annual Period 6 $2,955,010 $1,003,604 $2,336,169 $1,750,000 $8,044,783
la
Annual Period 7 $2,955,010 $1,003,604 $2,406,254 $1,750,000 $8,114,868
$2,955,010 $1,003,604 $2,478,442 $1,750,000 $8,187,056
ba
Annual Period 8
Annual Period 9 $2,955,010 $1,003,604 $2,552,795 $1,750,000 $8,261,409
$2,955,010 $1,003,604 $2,629,379 $1,750,000 $8,337,993
am
Annual Period 10
Annual Period 11 $2,955,010 $1,003,604 $2,708,260 $1,750,000 $8,416,874
Annual Period 12 $2,955,010 $1,003,604 $2,789,508 $1,750,000 $8,498,122
nj
Annual Period 13 $2,955,010 $1,003,604 $2,873,193 $1,750,000 $8,581,807

so
Annual Period 14 $2,955,010 $1,003,604 $2,959,389 $1,750,000 $8,668,003

Annual Period 15 $2,955,010 $1,003,604 $3,048,171 $1,750,000 $8,756,785
ck
TOTALS $43,476,313 $15,555,862 $37,984,280 $26,250,000 $123,266,455

Ja
As shown in Table 2, Siemens guarantees that the City will realize savings in four areas. First, Siemens
guarantees between $484,347 and $2,955,010 in additional annual revenue as a result of the greater
accuracy of the small meters installed. Second, Siemens guarantees between $501,802 and $1,003,604
in additional annual revenue as a result of the greater accuracy of the large meters. Third, Siemens
guarantees between $503,750 and $3,048,171 in annual operational savings related to meter reading
and billing as a result of the AMI Upgrades. Fourth, Siemens guarantees $1,750,000 in annual savings as
a result of being able to defer maintenance at the WTPs and the SCS. Over the term of the contract,
Siemens guarantees total savings of $123,266,455.
DETERMINATION OF ACTUAL SAVINGS
At the end of each contract year, Siemens will prepare an Annual Performance Assurance Report that
details the savings realized during the prior year and compares those saving to the guaranteed savings
Draft Final TM4-2

for that same period; however, only one category of savings is subject to annual verification in that the
savings resulting from large meter installation, the operational savings, and the deferred maintenance
savings are stipulated to have occurred each year regardless of whether any actual savings can be
demonstrated to have occurred. Therefore, of the approximately $123 million in anticipated savings,
only $43 million is actually guaranteed by Siemens while almost $80 million will only be realized as a
result of actions taken by the City.
The savings attributable to the small meter component of the AMS Upgrades is determined by
comparing the predicted billable water consumption using the existing meters for each meter group to
the billable consumption for each meter group using the new meters with the latter value being a
function of the measured accuracy of the new meters.
The calculation to determine the savings attributable to the new small meters is as follows:
om
 Predicted Existing Consumption (CCF) = Existing Consumption at 100% Accuracy
multiplied by the Predicted Meter Accuracy from Test Data applied per Guarantee Year
 Consumption Billed with New Meters (CCF) = Existing Consumption at 100% Accuracy x
.c
Actual New Meter Accuracy (%), where New Meter Accuracy is based on the
ya
measured and verified meter accuracy per Guarantee Year;
 Annual Consumption Increase (CCF) = Consumption Billed with New Meters (CCF) –
la
Predicted Existing Consumption (CCF); and
ba
 Annual Consumption Increase ($) = Annual Consumption Increase (CCF) x Water Rate
($/Unit).1
am
If the actual savings are less than the guaranteed savings, a Savings Shortfall will be paid by Siemens to
the City. The Savings Shortfall is an amount equal to the difference between the actual savings and the
nj
guaranteed savings.
so
It should be noted that the values for Existing Consumption at 100% Accuracy and Predicted Meter
ck
Accuracy are contractually predetermined, and therefore, the annual baseline consumption for each
meter group during each contract year can be calculated in advance and is shown in Table 3 on the
Ja
following page. As a result, the only variable used in the calculation of savings from the new meters is
the accuracy of the new meters which is determined using a testing protocol outlined in Exhibit C of the
contract.
1
The Project Team has developed an Excel spreadsheet that will allow the City to perform the calculation
described above.
Draft Final TM4-3

Table 3 – Baseline Consumption
Meter Groups Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7 Year 8 Year 9 Year 10 Year 11 Year 12 Year 13 Year 14 Year 15
Water - 5/8" Inside Group 1 1,227,323 1,220,278 1,213,233 1,206,187 1,199,142 1,192,096 1,185,051 1,178,005 1,170,960 1,163,914 1,156,869 1,149,823 1,142,778 1,135,732 1,128,687
Water - 5/8" Inside Group 2 1,089,119 1,082,867 1,076,615 1,070,363 1,064,111 1,057,859 1,051,607 1,045,354 1,039,102 1,032,850 1,026,598 1,020,346 1,014,094 1,007,842 1,001,590
Water - 5/8" Inside Group 3 753,675 749,349 745,022 740,696 736,369 732,043 727,716 723,390 719,063 714,737 710,410 706,084 701,757 697,431 693,104
Water - 5/8" Inside Group 4 893,832 888,700 883,569 878,438 873,307 868,176 863,045 857,914 852,783 847,652 842,521 837,390 832,259 827,128 821,997
Water - 5/8" Inside 1 Mile 236,762 235,403 234,044 232,685 231,326 229,966 228,607 227,248 225,889 224,530 223,171 221,812 220,453 219,093 217,734
Water - 5/8" Outside 1 Mile 185,806 184,739 183,673 182,606 181,540 180,473 179,406 178,340 177,273 176,206 175,140 174,073 173,007 171,940 170,873
om
Water -1" Inside 441,804 439,439 437,074 434,709 432,344 429,979 427,614 425,249 422,883 420,518 418,153 415,788 413,423 411,058 408,693
Water -1" Inside 1 Mile 9,018 8,969 8,921 8,873 8,825 8,776 8,728 8,680 8,632 8,583 8,535 8,487 8,438 8,390 8,342
Water -1" Outside 1 Mile 7,795 7,753 7,712 7,670 7,628 7,587 7,545 7,503 7,461 7,420 7,378 7,336 7,294 7,253 7,211
.c
Water -1.5"&2" Inside 1,574,867 1,566,552 1,558,237 1,549,922 1,541,607 1,533,292 1,524,977 1,516,661 1,508,346 1,500,031 1,491,716 1,483,401 1,475,086 1,466,771 1,458,456
Water -1.5"&2" Inside 1 Mile 8,293 8,249 8,205 8,162 8,118 8,074 8,030 7,986 7,943 7,899 7,855 7,811 7,767 7,724 7,680
ya
Water -1.5"&2" Outside 1 Mile 11,051 10,992 10,934 10,876 10,817 10,759 10,700 10,642 10,584 10,525 10,467 10,409 10,350 10,292 10,234
Sewer - 5/8" Inside Group 1 619,678 616,121 612,564 609,006 605,449 601,892 598,334 594,777 591,220 587,663 584,105 580,548 576,991 573,434 569,876
la
ba
Sewer - 5/8" Inside 1 Mile 5,505 5,473 5,442 5,410 5,378 5,347 5,315 5,284 5,252 5,220 5,189 5,157 5,126 5,094 5,062
Sewer - 5/8" Outside 1 Mile 845 840 835 830 825 821 816 811 806 801 796 792 787 782 777
am
Sewer -1" Inside 182,300 181,324 180,348 179,372 178,396 177,420 176,445 175,469 174,493 173,517 172,541 171,565 170,589 169,613 168,637
Sewer -1" Inside 1 Mile 772 768 764 760 756 752 748 743 739 735 731 727 723 719 715
Sewer -1" Outside 1 Mile - - - - - - - - - - - - - - -
Sewer -1.5"&2" Inside 1,132,801 1,126,820 1,120,839 1,114,858 1,108,877 1,102,896 1,096,915 1,090,934 1,084,953 1,078,972 1,072,991 1,067,010 1,061,029 1,055,048 1,049,067
Sewer -1.5"&2" Outside 1 Mile 894 889 885 880
nj 875 870 866 861 856 851 847 842 837 833 828
Sewer -1.5"&2" Inside 1 Mile - - - - - - - - - - - - - - -
so
Total 9,903,039 9,847,695 9,792,352 9,737,008 9,681,665 9,626,321 9,570,978 9,515,634 9,460,291 9,404,947 9,349,604 9,294,260 9,238,917 9,183,573 9,128,230
ck
Ja
Draft Final TM4-4

Since each of the other three components of annual savings are stipulated in the contract and the
savings attributable to new small meters is solely a function of the measured accuracy of the new
meters, it will be very important for the City to carefully monitor the testing of the new meters.
METER TESTING PROTOCOL
Article 4 of Exhibit C of the contract describes the testing protocol that will be used each year to
determine the accuracy of the new meters. Generally speaking, Siemens will select a sample of meters
using a process that randomly selects a predetermined number of meters from each meter group. The
contract stipulates that the sample from each meter group will include meters that have experienced
high, medium, and low cumulative flows over their lifetime. The inclusion in the testing sample of
meters that have experienced various levels of flow recognizes that meter accuracy decreases more
om
quickly in meters with high flow levels as a result of wear on the meter.
Additionally, as shown in Table 4 below, the contract stipulates the sample size for each meter group
.c
based on the number of meters in each group.
Table 4 – Meter Sample Size Guidelines

ya
la
ba
Population Number of Samples
1 1
am
2 2
3 3
nj
4 3
5-6 4
so
7-9 5
10-13 6
ck
14-19 7
Ja
20-29 8
30-49 9
50-110 10
>110 11
Sample sizes for each meter group using the criteria from Table 4 are provided in Table 5 on the
following page. The largest group of meters, 5/8” Inside meters, were broken into four groups, and
these groups were utilized for determining the sample size.
Draft Final TM4-5

Table 5 – Meter Testing Sample Sizes2
Meter Groups Total Meters Sample Size

Water - 5/8" Inside Group 1 11,121 11
Water - 5/8" Inside 1 Mile 1,911 11
Water - 5/8" Outside 1 Mile 3,587 11
Water -1" Inside 2,091 11
Water -1" Inside 1 Mile 35 9
Water -1" Outside 1 Mile 100 10
Water -1.5"&2" Inside 1,866 11
om
Water -1.5"&2" Inside 1 Mile 19 7
Water -1.5"&2" Outside 1 Mile 19 7
Sewer - 5/8" Inside Group 1 10,392 11
.c
ya
Sewer - 5/8" Inside 1 Mile 112 11
la
Sewer - 5/8" Outside 1 Mile 11 6
ba
Sewer -1" Inside 1,415 11
Sewer -1" Inside 1 Mile 9 5
am
Sewer -1" Outside 1 Mile - 0

Sewer -1.5"&2" Inside 1,368 11
Sewer -1.5"&2" Inside 1 Mile 5 4
nj
Sewer -1.5"&2" Outside 1 Mile - 0

so
Total Meters To Be Tested 213

ck
Once the sample meters have been selected, they will be removed from service and tested using
procedures provided in the American Water Works Association (AWWA) Manual M6 – Water Meters-
Ja
Selection, Installation, Testing, and Maintenance. The test for each meter will involve the determination
of meter accuracy at three flow levels (low, medium and high) and the average weighted efficiency of
the meter will be determined. Once all of the sample meters from a specific Meter Group have been
tested, the results will be averaged to determine the accuracy for that Meter Group. As described in
Article 4 of the contract, this value will be multiplied by the value for Existing Consumption at 100%
Accuracy for that meter group to arrive at the Consumption Billed with New Meters for the Meter Group
which is then used to determine whether or not the Performance Guarantee has been met for the
contract year.
2
Meter counts for each meter group are based on data extracted from the City’s water billing system in November
2014. The actual number of meters tested will need to be determined annually based on the number of meters
actually in service at the time.
Draft Final TM4-6

Section 4.6.1 of the contract discusses the possibility of Siemens taking corrective action in the event
that a Savings Shortfall occurs, but since measured accuracy of the new meters is the only factor that
can affect whether or not a Savings Shortfall occurs, it is unlikely that Siemens could take any action that
would remedy a Savings Shortfall that has already occurred. Therefore, if a Savings Shortfall does occur,
Siemens’ only recourse would be to pay the City an amount equal to the Savings Shortfall.
OTHER COMPONENTS OF GUARANTEED SAVINGS
As mentioned previously, the Operational Savings, Deferred Maintenance Savings and additional
revenue resulting from the installation of large meters are contractually considered to be Stipulated
Savings which means they are presumed to have occurred regardless of whether the City actually
realizes savings in these areas. As such these savings have no impact on whether or not Siemens meets
its Guaranteed Savings. However, these potential savings are to one degree or another attainable and
the City should strive to ensure these savings are actually realized.
om
Operational Savings
.c
Siemens maintains that the City will realize operational savings as a result of several factors associated
with the new AMI meters.

ya
Fewer New Employee Requests – The City currently uses manual labor for meter reading and
la
meter data entry. The remote read capabilities of the new meters will result in the need for
ba
fewer employees and the City should realize a commensurate decline in employee costs.
 Reduced Vehicle Costs – Meter readers currently use vehicles to travel to meter reading routes.
am
Since meter readings will be taken remotely, meter readers will no longer need to travel to
meter reading routes which means they will not need vehicles for travel. As a result the City will
avoid the cost of vehicles and vehicle fuel.
nj
 Fewer Re-reads/Meter Locates – Currently, when it appears that a reading error may have
so
occurred, a meter reader must return to the meter location to perform a re-read. Additionally,
meter readers sometime have difficulty locating meters and the utility incurs costs to locate the
ck
meters such that they can be read. The remote read capabilities should reduce these costs in
several ways. First, the remote reads are typically more accurate than manual reads thereby
Ja
reducing the number of re-reads necessitated by reading errors. Second, in a case where a re-
read is required, the re-read can be performed remotely thereby eliminating the need for a
person to travel to and read the meter. Third, since remote reads occur regardless of whether
someone knows the exact location of the meter the need for meter locates should be reduced.
 Fewer Meter Lid Replacements – The new meters will be fitted with tighter fitting, non-floating
lids that should stay in place better than the current meter lids. This will reduce the number of
meter lids that require replacement each year.
 More Efficient Shut offs/Service Restorations – Currently, when service to a location needs to be
discontinued or restored a person must travel to the location and physically turn the water
service off or on at the meter. The new meters have remote turn of/turn on capability thereby
reducing, if not eliminating, the need for a person to travel to the location.
Draft Final TM4-7

 Reduced Bill System Maintenance – The existing billing system requires approximately $20,000
in maintenance each year. The new billing system will also require annual maintenance, but
these costs are included in the amount paid to Siemens under the contract.
Table 6 below summarizes the operational savings that Siemens maintains the City will realize as a result
of the metering and billing system project.
Table 6 – Anticipated Operational Savings
Annual Operational Savings Calculations

Quantity Source Unit Total Annual Total
14 New Employee Request Reduction (Annual Cost) $34,000 $476,000
11 Vehicle Reductions (Annual Cost - $350/month/vehicle) $4,200 $46,200
om
11 Vehicle Fuel (Average Cost) $3,000 $33,000
20,000 Annual Re-reads/Meter Locates (100 per day * 200 days/yr) $20 $400,000
3,000 Meters/Lids Replaced Annually ($80 per meter) $80 $240,000
.c
20,000 Annual Shut offs/Restore service (100/day * 200 days/yr) $40 $800,000
ya
1 Existing Billing System Maintenance
la $20,000 $20,000
It should be noted that the savings associated with Re-reads/Meter Locates; Meters/Lids Replaced;
ba
Annual Shut offs/Restore service; and Existing Billing System Maintenance are predicated on the
assumption that the City currently spends an amount that is at least equal to the predicted savings
am
associated with each of these items. If the City actually spends less than the predicted amount, then
realized savings will be equal to the amount actually spent and not the amount predicted by Siemens.
For example, if the City currently only performs 90 Re-reads/Meter Locates per day, then the City will
nj
only realize savings associated with 90 Re-reads/Meter Locates instead of the 100 used to estimate
so
savings. In order to set reasonable expectations for the level of operational savings that will be realized
in association with each of these items, the City should determine the amounts they are currently
ck
spending on each.
Ja
While it is possible that the City will realize savings in each of these areas it is imperative that the City
take action to maximize the savings it will ultimately realize. First, the City should make sure that the
new meters are installed as quickly as possible. All of the operational savings presented in Table 6 are
contingent upon the entire system being equipped with remote read meters. Until all of the new
meters are installed there will still be a need to maintain the staff necessary to manually read and locate
meters and to discontinue and restore service.
Second, the City must effect the staff reductions in order to realize the anticipated savings. If staff count
is not reduced these savings will not be realized. It should be noted that based on our experience with
water utilities across the country, municipal labor policies often make it very difficult to effectuate staff
reductions in a timely manner. Typically staff reductions are only realized as the result of attrition and
therefore decreased costs as a result of staff reductions may not be realized as quickly as the contract
predicts. In addition, the AMI system will still require annual inspections of meters and billing personnel
Draft Final TM4-8

will be needed to address the billing and collection issues noted in Task 3. City staff also indicated a
team of existing customer service representatives will be needed to test the system. If these employees
are testing the system, other employees will be needed to fill their positions.
Deferred Maintenance Savings

Siemens maintains that the City will be able to defer collection system maintenance costs as a result of
the collection system improvements included in the Siemens contract. Since many of these projects
involve replacing older, more maintenance intensive assets with newer assets, it is possible that
maintenance costs will be avoided in the short-term.
Large Meter Savings

Siemens is replacing a number of large meters with new remote read meters. The “savings” attributable
to the large meter replacements is actually additional revenue that will accrue to the City as a result of
greater metering accuracy. These savings are similar to the savings attributable to the small meter
om
replacement program; however, unlike the small meter savings, this component of saving is not based
on testing of the actual accuracy of new meters. Instead, these savings are Stipulated Savings and will
.c
be assumed to have occurred regardless of whether or not more large meter revenue is generated.
Since it is likely that the new meters will be more accurate than the meters they are replacing, the
ya
probability of realizing at least a portion of these savings is high; however, there is no guarantee that the
additional revenues will be equal to or greater than the stipulated savings included in the performance
la
guarantee. It is not clear why the contract does not treat large meters in a manner similar to the
ba
treatment of small meters and provide a guarantee of additional revenue based on actual large meter
accuracy, but since it does not, in order to maximize the additional revenue realized from these large
am
meters the City must ensure that these meters are indeed more accurate than the meters they replaced.
To do this, the City must develop and implement a robust meter testing and maintenance program for
nj
these meters.
so
STATUS OF CONTRACT PROJECTS

ck
As of October 2014, Siemens had invoiced the City a total of $74,522,254 which represents
Ja
approximately 82% of the contract value. As of that same date the WTP projects were approximately
76% complete and the SCS projects were approximately 89% complete.
Although the total amount invoiced includes charges for approximately 89% of the new meters, it is our
understanding that only around 40% of the new small meters had been installed as of the end of
October. Additionally, Siemens has charged the City for 79% of the costs associated with the new billing
system.
CONCLUSIONS AND RECOMMENDATIONS
The projects being implemented under the Siemens contract should allow the City’s water and sewer
utilities to improve their efficiency, especially in the areas of meter reading and billing. Specifically, the
installation of remote read meters and a new billing system should result in increased revenue by
Draft Final TM4-9

ensuring that water bills more accurately reflect the amount of water being consumed. In fact, Siemens
has guaranteed a certain level of increased revenue associated with the meters that are typically used to
serve residential customers. However, the majority of the “guaranteed savings” that Siemens maintains
will be realized during the contract term are not truly guaranteed. Only the additional revenue that the
City should realize as a result of the greater accuracy of the new meters 2-inches and smaller is
guaranteed and in reality, only the accuracy of those meters is guaranteed.
The other projects implemented under the Siemens contract are expected to either reduce expenses
associated with meter reading, billing and maintenance or provide additional revenue as a result of
greater accuracy of new meters larger than 2-inches.
By taking the following steps, the City can improve the probability that the benefits offered by the
Siemens contract are realized.
om
 Prioritize the installation of the remote read meters. Until such time as the entire system is
equipped with remote read meters the accuracy of the new meters cannot be determined and
.c
the City will not realize the savings that are dependent upon reduced staffing requirements.
 Since it is quite possible that complications associated with integration of data derived from the
ya
new, remote read meters and data derived from old, manual read meters will reduce efficiency
la
and lead to greater costs, the City should postpone implementation of the new billing system
until all of the remote read meters are installed.
ba
 While the full implementation of the new billing system should not take place until all of the
remote read meters are installed, the City should work with Siemens to ensure that all actions
am
required to allow for the implementation of the new system are being performed. To the extent
that these actions require City resources, the City should do everything in its power to ensure
nj
these resources are available. There are a couple of other tasks the City should undertake prior
to implementation.
so
1. Process map existing billing and collection practices and processes under the new
billing/AMI system
ck
2. Modify policies (adjustments, cut-offs, etc.) accordingly

Ja
 Once the system is fully equipped with remote read meters, the City must effect the staff
reductions that drive the operational savings. If current staffing levels are maintained, these
operational savings will not be realized.
 Determine the amount that is currently being spent for the items included under Operational
Savings such that the City will have more accurate expectations for the savings that will be
realized once these expenditures are either reduced or eliminated as a result of the installation
of the remote read meters.
 Carefully monitor the meter testing program that determines whether the City is actually
generating more revenue as a result of the greater accuracy on the small meters. Although the
meter testing protocol described in the contract should provide a good indication of the
accuracy of the new small meters, the City must ensure that the test sample is truly
representative of the meters throughout the system.
Draft Final TM4-10

 Develop a robust testing protocol for the new large meters that ensures that these meters are
providing the expected benefits. While the additional revenue from large meters is not
guaranteed, it is still important to make sure that the new large meters are accurately recording
water consumption.
om
.c
ya
la
ba
am
nj
so
ck
Ja
Draft Final TM4-11

Siemens Raftelis Phase 1 Report WM - Draft Final

Uploaded by

Copyright:

Available Formats

Siemens Raftelis Phase 1 Report WM - Draft Final

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Siemens Raftelis Phase 1 Report WM - Draft Final

Uploaded by

Copyright:

Available Formats

1031 S. Caldwell Street Phone 704 . 373 . 1199 www.raftelis.

February 27, 2015

Ms. Kishia Powell

Re: Phase 1 Final Report – Draft Final

 Task 2 – Revenue Sufficiency Analysis

 Task 4 – Siemens Contract Review

 Task 6 – Additional Efforts

TASK 1 –STRATEGIC FINANCIAL PLANNING

TASK 2 – REVENUE SUFFICIENCY ANALYSIS

Draft Final ES-1

TASK 3 – BILLING SYSTEM DATA REVIEW

stored in the legacy billing database.

Draft Final ES-2

TASK 4 –SIEMENS CONTRACT REVIEW

Draft Final ES-3

associated with the Siemens’ team performance under the contract.

TASK 5 – INFRASTRUCTURE MASTER PLAN DEVELOPMENT

Draft Final ES-4

TASK 6 – ADDITIONAL EFFORTS

Draft Final ES-5

Draft Final TM1-1

Draft Final TM1-2

Next permit renewal is in 2017. Addressing

Addressing this might hinge on determining

City needs to determine refunds to those

Consider conducting a water loss audit to

Draft Final TM1-A1

Review current accounting practice

Cost of effort (staff time and

Draft Final TM1-A2

Included in current scope of work with RFC

Need to finalize Infrastructure Master

Goal should be to identify who the

Identify universe of contracts that should

be assessed. Need to set a limit as to how

Draft Final TM1-A3

Chemicals and electricity are typically the

Cost of effort (staff and consultant

Cost of discussions (staff and

Cost of discussions (staff and

Draft Final TM1-A4

Draft Final TM2-1

Revenue Requirements Assumptions

capital projects, which are explained below:

anticipated future costs.

Draft Final TM2-2

stronger debt service coverage. RFC is recommending an internal target of cash

Draft Final TM2-3

Debt Service Coverage Calculation Assumptions

otherwise provided in this subsection

which is incorporated into the financial planning model.

Financial Plan Results

Draft Final TM2-4

contract resulting from increased meter accuracy.

given to customers, enforcing the cut-off policy, etc.

Draft Final TM2-5

COMPARISON OF STUDY RESULTS TO ENGINEER’S PROJECTIONS

Draft Final TM2-6

Timing Event Impact