ETL 1110-2-550
30 May 97
estimated R(t), the hazard function can be derived
(Russell et al. 1995). The HYD-QUAD
and the failure rate as a function of age can be
development is based on the NAV-QUAD work
found.
(Russell et al. 1993). The purpose of the HYD-
QUAD is to develop a program that can be used to
(3) The methodology used to conduct this part
assist Corps managers in allocating maintenance
of the study was excellent but the results are
funds. The project develops and uses an analytical
expected to have larger uncertainties associated
process based on economics to make maintenance
with them. Additionally, the method is generally
investment decisions. HYD-QUAD is generally
more difficult to apply especially when considering
applicable to maintenance items that are not a
the age effects of random variables. A capacity and
baseline budget item, have a cost greater than
demand formulation could be used, but fitting
0,000 and less than ,000,000, and have an
historical survivor data with a probility density
impact on the probability of plant outage.
function (e.g., Weibull or log-normal distribution) is
b. For a given project or facility, the HYD-
preferable.
QUAD framework evaluates an initial maintenance
condition, impact on maintenance condition, and
F-5. Turbine Reliability
leads to a maintenance project ranked on economic
impact. The HYD-QUAD model was developed
JAYCOR prepared this report which documents the
using the existing NAV-QUAD model, a literature
development of quantitative measures for the
review, interviews with field professionals,
reliability of turbine features using capacity and
workshops, and focus group meetings. Generally,
demand analysis (Mlaker and Bryant 1994). The
the HYD-QUAD methodology was developed
results can be used in economic models which
around the following four steps:
optimally allocate limited resources for project
rehabilitation. The second report section introduces
Step 1 - Determine the current condition of the
general approaches to reliability estimation that are
facility for which maintenance is being considered
used in the study. In the third section, deterministic
using the CI method developed under the REMR
models for three modes of unsatisfactory perform-
program (Norlin et al. 1993). The CIs for each
ance of the turbine hub are described. In section
component/system are combined to produce an
four, a probabilistic-based, reliability formulation is
overall indicator of the facility condition, called a
explained for application to these three models of
summary condition index (SI). This SI value is
unsatisfactory performance. The resulting
obtained by taking a weighted average of each of
probabilistic models are then applied to turbine unit
the components that comprise the unit or facility.
number 3 at the Walter F. George Power Plant. In
These weighting factors were determined by
the fifth and closing section of the report,
averaging 38 estimates made by hydroelectric
observations about the reliability of these features
power managers. Each manager was asked to
of hydropower equipment are summarized and
distribute 100 points among the components/
recommendations are made for broadly applying the
systems of the facility or unit. Hydroelectric
method to Corps hydropower projects.
equipment included in the initial evaluation are
generator, transformer, circuit breaker, governor,
and voltage regulator.
F-6. Hydroelectric Power Quadrant
Prototype
Step 2 - Each proposed maintenance alternative
is listed and the associated change in SI is estimated
a. The QUADRANT model for hydroelectric
to produce a "change in SI." Information on current
unit or facility condition and the expected improve-
power (HYD-QUAD) is being developed
ment on condition is provided by districts or
specifically for field use on nonroutine main-
projects.
tenance studies to support project budget decisions
F-4
Previous Page
