Due to the increase of using DG in power networks and the diversity in renewable energy resources, new optimal protection relay settings need to be considered in order to keep power system stability [1] .

Some types of DG can be used according to loading conditions, it can be used at summer only or at rush hours, also some types can work at daylight like PV systems; therefore, protection system should be adaptive to generation condition and should be flexible to system status.

When a micro-grid is connected to power network, the configuration is changed to a complicated multi-source power system, fault currents and fault levels need to be recalculated and power ratings of protection equipment’s should be rearranged accordingly, the protection of microgrid should be in such a way that a safe and secure protection is provided in both grid-connected and stand-alone operation modes [2] [3] .

Modern multifunction digital relays have a number of features, which make them an ideal choice for interconnection protection of dispersed generators. The most important of these features are user-selectable functionality, self-diagnostics, communications capabilities and Oscillo graphic monitoring [4] .

A multiple setting group protection relay has the ability to change setting locally with selector or remotely through communication system, as it has the ability to move between up to eight different setting groups [5] .

Central protection units communicate with every network component such as breaker status and relay setting and each new connection/disconnection is reported to it. Therefore, Central control unites have the ability to extract the current state of the network, list the connected entities and choose a proper relay setting [6] .

The optimum settings of DOCRs need to be determined for all groups. Here, the major task is to obtain values of TMS of DOCRs suitable for proper protection coordination under all the topologies considered in each group. This can be done by considering all the primary-backup relay pairs of all the topologies of a group so that the obtained optimum values of TMS can coordinate properly in these topologies [7] .

Microgrid Central Protection Unit (MCPU) communicates with every single relay and distributed generator in the microgrid. The communication with relays is necessary to update the operating currents of the relays and to detect the direction of fault currents and thus isolate the fault properly. DGs, on the other hand, are monitored to follow their status and include/disregard their fault contribution if they are ON/OFF, respectively [3] .

MCPU is used in some studies to be responsible of calculation of tripping currents, time delays (optional) and updating of relay operating point [8] and [2] , On the other side, in another study, the programmable scheme logic in modern multifunctional protective relays is used, it’s an extremely powerful tool that allows the user to adapt the relay logic to very different applications or to change system conditions [9] , while in our study all tripping currents and time delay at different operating cases will be stored in relay at different setting group, MCPU will choose optimal setting group needed according to generation capacity.

Our proposed MCPU shown in Figure 1 doesn’t need to be updated with system current and voltage, just it will supervise if any change happens in circuit breakers status, if any change is noticed in system, MCPU will send signal to all DOCRs and select suitable setting group to be activated in such case.

With the wide use of optic fiber communication in distribution grid, it is practicable to achieve MCPU logic for microgrid [10] .

One of the most important advantages in our proposed MCPU, all relays will be restored to default setting group, in case of communication fail between MCPU and relays or breakers and in case of any fault happed, relay will send trip signal to breaker.

In order to keep the system stable under any fault condition, protection relay should isolate faulted area in the instance of fault occur, if this relay fails to operate under any condition, nearest relay to faulted zone should operate as a backup relay.

Earlier studies in directional over current relay (DOCR) coordination was using protection relay fixed with single setting group [11] .

For multiple setting protection relay, the function of relay to act as primary or backup depends on the direction of fault current, each multiple setting DOCR has many sets of independent relay settings we can use one for forward direction as a primary relay and another one in reverse direction of fault current as backup relay [12] .

The difference in coordination between the primary back-up relay pairs for proposed and conventional back up DOCR is shown as in Figure 2 .

Figure 2 shows apart of IEEE 8-bus system, in case of occurrence of a fault at the point f1, the relay (i) acts as primary relay which should operate to isolate the fault, if this relay fails to operate and after a certain time, another relay should operate as a backup relay this backup relay will be as follow:

Therefore, the reverse side of each relay is deployed as the backup relay for the next front line where in our case according to fault direction relay (k) will select setting group for forward fault current relay will work according to setting group (1) and reverse direction fault current the relay will work according to setting group (2).

To avoid the operation of relay (j), a communication link is considered to send a block signal on behalf of relay (k) reverse [13] .

The operating time of inverse definite minimum time relay IDMT is obtained according to IEEE standard C37.112-1996: [14]

$t_i=TD{S}_i\times \left(\frac{28.2}{{\left(\frac{I_{fi}}{I_{pi}}\right)}^2-1}+0.1217\right)$(1)

where

The value of the pickup current of a relay (i) should be less than the minimum value of the fault current at this point [15] .

Furthermore, the time dial setting TDS_i of relay i is ranged from 0.1000 to 1.1000 [16] .

CTI value ranges from 0.2 to 0.3 sec. In this paper 0.2 sec value is chosen [17] .

In case of using conventional DOCR our object function will be:

$objectfunction\left(J\right)={\displaystyle {\sum }_{i=1}^n{t}_i}$ [18]

In this case we need to get the optimal values of pickup current and TDS values of each relay.

While in case of using proposed multiple setting DOCR, each relay will be installed with two different setting , first one will be used at forward fault and second setting group will be used in reverse direction faults and object function will be as equation:

$objectfunction\left(J\right)={\displaystyle \underset{i=1}{\overset{n}{\sum }}\left(t_{\left(iforward\right)}+t_{\left(ireverse\right)}\right)}$

In this case the parameters which need to be optimized for each relay will be duplicated, where two values of pickup current and TDS is needed to each relay.

To study the effect of using multiple setting DOCR in micro-grid systems IEEE 8-bus test system is used, and in order to clarify our study, two different types of distributed generator (DG) is connected at bus-2, first is photo voltaic (PV) system and second is wind turbine generator (WTG) as shown in Figure 3 .

All data of our studied IEEE 8-bus test system which shown in Figure 4 taken from [19] , the data for proposed DGs available at [20] .

To get the value of maximum short circuit current sensed at each relay on the system, system is simulated using Etap software as shown in Figure 5 , for simplicity the values of pickup current of each relay is considered constant and taken from [14] .

Grey Wolf Optimizer (GWO) [21] algorithm is developed using MATLAB software to achieve optimal TDS values.

IEEE 8-bus system is simulated using Etap software, our study will be concentrated only under three different cases shown in Table 1 and Figure 3 .

Case-1 CB-15 is opened, which means no DGs is connected at bus-2.

Case-2 CB-15 is closed, CB-16 is closed and CB-17 is opened, in this case PV system connected at bus-2.

Case-3 CB-15 is closed, CB-16 is opened and CB-17 is closed, in this case WTG system connected at bus-2.

For the three different cases the maximum fault current sensed by each relay is calculated with ETAP and the results is shown in Table 2 , moreover the conventional and proposed backup relays also considered in study as in Table 3 .

GWO algorithm used to get optimal TDS value for each relay at different cases and the result TDS values in case of using conventional backup relays is shown in Table 4 , while in case of using proposed backup relay, every relay will be installed with two different TDS per case, which means every relay will be installed with 6 six different setting groups as in Table 5 , while in conventional backup relay only three setting group needed per relay because same value of TDS will be used in case of primary and backup protection.

Using the TDS values from Table 4 and Table 5 , the value tripping time for primary and backup relays is calculated in case of conventional and proposed backup relays at the three different cases and shown in Table 6 and Table 7 .

From Table 6 , it can be noted that primary relays tripping time in case of using proposed backup relay is less than its values when using conventional backup relay, as in Table 6 trip time is reduced by 33, 35 and 37% for case-1, case-2 and case 3 respectively, which increases system performance in decreasing total tripping time.

Now, after getting the optimal TDS values for all relays, every relay will be programmed with six setting groups in case of using proposed backup relay method, or three different setting groups in case of using conventional backup relay method.

MCPU in Figure 1 is used to supervise the CBs condition as in table I, and upon this condition, MCPU will chose suitable setting group, and make it active through communication link with all relays, the proposed protection system with logical nodes and data sets can be provided by IEC 61850-7-420 [12] .

Figure 6 shows flow chart of our study in case of using proposed backup relay method, MCPU will keep connected with system until detecting any change of CB15, CB16 and CB17, and upon this change it will change setting group according to Table 5 .

This paper presented effect of using MCPU in microgrid, in case of connecting DGs, MCPU monitor system status and changes relay setting group according to generation capacity, while in case of taking fault current direction in our consideration, a proposed backup relay will be useful in decreasing relays tripping time.

ETAP software is used to evaluate fault current which sensed at each relay, while MATLAB software is used to implement GWO algorithm which is used to get optimal TDS value for each relay.

The value of CTI between primary and backup relay in each case is found to be more than 0.2 sec, which achieves selectivity in our proposed protection system.

It’s found that MCPU is important in case of presence DGs in system, MCPU will keep relay coordination and protection stability at right way even if changing fault current due to DGs connecting in microgrid.

Our proposed MCPU was designed with simple rules which doesn’t require a special system memory or data processing unite, sense it depends only on a simple logic in changing relay setting group according to DGs connection in microgrid.